Synthesis of Discipline-Based Education Research in Physics

PHYSICAL REVIEW SPECIAL TOPICS - PHYSICS EDUCATION RESEARCH 10, 020119 (2014)
Synthesis of discipline-based education research in physics

Jennifer L. Docktor1 and Jos P. Mestre2,*
1
Department of Physics, University of Wisconsin-La Crosse, 1725 State Street,

La Crosse, Wisconsin 54601, USA
2
Department of Physics, University of Illinois, 1110 West Green Street, Urbana, Illinois 61801, USA
and Department of Educational Psychology, University of Illinois,
1310 South 6th Street, Champaign, Illinois 61820, USA
(Received 8 August 2013; published 16 September 2014)
This paper presents a comprehensive synthesis of physics education research at the undergraduate level.
It is based on work originally commissioned by the National Academies. Six topical areas are covered:
(1) conceptual understanding, (2) problem solving, (3) curriculum and instruction, (4) assessment,
(5) cognitive psychology, and (6) attitudes and beliefs about teaching and learning. Each topical section
includes sample research questions, theoretical frameworks, common research methodologies, a summary
of key findings, strengths and limitations of the research, and areas for future study. Supplemental material
proposes promising future directions in physics education research.
DOI: 10.1103/PhysRevSTPER.10.020119
PACS numbers: 01.40.Ha, 01.40.Fk, 01.40.gb
I. INTRODUCTION
This paper synthesizes physics education research (PER)
at the undergraduate level, and is based on a paper that was
commissioned by the National Research Council to inform
a study on the status, contributions, and future directions
of discipline-based education research (DBER)a comprehensive examination of the research on learning and
teaching in physics and astronomy, the biological sciences,
chemistry, engineering, and the geosciences at the undergraduate level [1,2]. PER is a relatively new field that is
about 40 years old, yet it is relatively more mature than its
sister fields in biology, chemistry, engineering, astronomy,
and geosciences education research. Although much is
known about physics teaching and learning, much remains
to be learned. This paper discusses some of what the PER
field has come to understand about learners, learning, and
instruction in six general topical areas described herein.
A. Topical areas covered and organization
Given the breadth and scope of PER to date, we organize
this synthesis around six topical areas that capture most of
the past research in physics education: conceptual understanding, problem solving, curriculum and instruction,
assessment, cognitive psychology, and attitudes and beliefs
about learning and teaching. To ensure consistency in the
presentation and to aid the DBER committee in its charge,
each of the six topical areas is organized under the
*
[email protected]
Published by the American Physical Society under the terms of

the Creative Commons Attribution 3.0 License. Further distribution of this work must maintain attribution to the author(s) and
the published articles title, journal citation, and DOI.
1554-9178=14=10(2)=020119(58)
following sections: research questions; theoretical framework; methodology, data collection or sources and data
analysis; findings; strengths and limitations; areas for
future studies; references. In this paper, the final section
on continuing and future directions of physics education
research has been removed and placed in the Supplemental
Material [3]. In addition, the references have been compiled
at the end of the paper rather than individually by section.
Because of the cross-cutting nature of some articles, some
that were included in a particular section could have just as
easily been included in another section; we highlight the
specific features of articles as they pertain to the sections
emphasis. Although we did not place any restrictions on the
dates of the research studies covered, the great majority of
the studies cited are within the past 20 years and were done
in the United States. The original commissioned paper
included published studies up through October of 2010,
and this revised paper includes studies published through
May of 2013. In addition to the six topical areas, a
summary and our conclusions are presented in Sec. VIII.
Equally important to stating what this paper is covering
is stating what has been left out. The commissioned paper
had a specific focus on empirical research on undergraduate teaching and learning in the sciences as outlined by the
criteria set by the National Academies [1,2]. Therefore, the
following areas of research have not been included in this
review: precollege physics education research (e.g.,
research on high school physics teaching and learning)
and research related to physics teacher preparation or
physics teacher curricula. We also made a decision to
exclude how to articles describing research analyses
(e.g., ways of analyzing video interviews of students)
which are pertinent for the community of physics education
researchers but do not have a direct impact on undergraduate physics education. The coverage herein is extensive,
020119-1
Published by the American Physical Society
JENNIFER L. DOCKTOR AND JOS P. MESTRE
PHYS. REV. ST PHYS. EDUC. RES 10, 020119 (2014)
although not exhaustive; whenever multiple articles were

available on the same or similar topics, we selected
representative articles rather than including all. For areas
of research that have been around longer in PER (e.g.,
conceptual understanding and problem solving), there exist
review articles (which we cite), thereby helping to reduce
the length of those sections. Emerging areas of PER, on the
other hand, lack review articles and hence those sections
may be slightly longer for adequate coverage.
It is somewhat surprising to us that this is the first PER
review article appearing in Physical Review Special
TopicsPhysics Education Research (PRST-PER), given
the large body of work that exists in PER and that review
articles are among the types of articles solicited in the
About description of the journal. Although we have
attempted to present a balanced view, it is unavoidable that
the contents herein will reflect to some extent the views and
biases of the authors. We hope to see more review articles in
PRST-PER in the future. We apologize if we have left out a
particular researchers favorite workif we did, it was not
intentional.
common student misconceptions [4,5], and there were

entire conferences devoted to student misconceptions in
the Science, Technology, Engineering and Mathematics
(STEM) disciplines [6,7] with thick proceedings emerging
from them. Many of these studies also included the
development of instructional strategies and curricula to
revise students thinking to be in alignment with appropriate scientific explanations, a process referred to as
conceptual change. Instructional strategies and curricula
are described Sec. IV, and the development of conceptual
exams is described in Sec. V.
Recent studies on students conceptual understanding
have broadened the focus from common difficulties to
generating theories and explanations on the nature and
origin of students ideas, and describing how those ideas
change over time.
A. Research questions
The research questions investigated under the conceptual
change generally fall into the following categories.
1. Identifying common misconceptions
B. Possible uses of this synthesis

We envision multiple uses of this synthesis. First, it
captures a good cross section of PER, and as such it is a
good resource for physics faculty, discipline-based education researchers, and, in particular, PER faculty, postdoctoral candidates, and graduate students; the contents herein
provide an excellent resource for those interested in an
overview of the PER field at the postsecondary level, and
would be useful in teaching a graduate seminar on PER.
Second, it serves as a historical account of the field, taking
stock in where PER has been and where it currently is.
Finally, it provides a perspective of the status of other
discipline-based education research relative to physics.
II. CONCEPTUAL UNDERSTANDING
One of the earliest and most widely studied areas in
physics education research is students conceptual understanding. Starting in the 1970s, as researchers and instructors became increasingly aware of the difficulties students
had in grasping fairly fundamental concepts in physics
(e.g., that contact forces do not exert forces at a distance;
that interacting bodies exert equal and opposite forces
on each other), investigations into the cause of those
difficulties became common. Over time these conceptual
difficulties have been given several labels, including
misconceptions, naive conceptions, and alternative conceptions. In this review we have chosen to use the term
misconceptions but acknowledge that some researchers
may have a preference for other terms.
Some would argue that misconceptions research marked
the beginning of modern-day physics education research.
Early work consisted of identifying and documenting
What learning difficulties do students possess that are

conceptual in nature? What are the most common misconceptions that interfere with the learning of scientific
concepts? Much work has gone into documenting preconceptions that students bring into physics classes prior to
instruction and identifying which of those are misconceptions that are in conflict with current scientific concepts
(see, e.g., Refs. [8,9]). Although many of these studies
address topics in mechanics (e.g., kinematics and dynamics), there have also been many studies conducted in
electricity and magnetism, light and optics, thermal physics, and a few in modern physics. For a list of approximately 115 studies related to misconceptions in physics,
see a resource letter [10].
In addition, many investigations explore whether or not
misconceptions persist following instruction, which in turn
provide insights into the type of instruction that impacts
students conceptual understanding. This body of work
has generated numerous carefully documented studies
(see, e.g., Refs. [11,12]) of the role of misconceptions in
students reasoning and learning, as well as several inventories to assess conceptual understanding in several physics
domains that will be discussed further in the Sec. V
[1317].
2. Describing the architecture of conceptual structure
What is the nature of scientific concepts in memory?
What changes when conceptual change takes place?
Another thrust in research on students conceptual understanding attempts to describe the cognitive architecture of
conceptual knowledge (i.e., how conceptual knowledge is
structured in memory). This body of work has generated
lively debate among those proposing different cognitive
020119-2
SYNTHESIS OF DISCIPLINE-BASED
architectures [1823], and some have combined different

types of cognitive architectures to explain student reasoning [24]. An interesting emerging line of work in misconceptions uses functional magnetic resonance imaging
(fMRI) to investigate the nature of conceptual change;
preliminary findings suggest that even when students know
the right answers (i.e., when they have supposedly overcome their misconceptions) brain activation suggests that
many students may still hold the misconception in memory
yet suppress it [25]. Understanding how concepts form in
memory and how they are used to reason provides useful
insights in devising instructional interventions to help
students adopt and use scientific concepts.
from these and multiple similar observations construct the

notion that heavy objects fall faster than light objects.
Misconceptions are viewed as stable entities that are used to
reason about similar but varied contexts (e.g., a sheet of
paper falling from a table to the floor falls slower than a set
of keys falling to the floor, both observations fitting into
heavy objects fall faster than light objects misconception). Misconceptions have three qualities: (a) they interfere
with scientific conceptions that teachers attempt to teach in
science classes, (b) they are deeply seated due to the time
and effort that students spent constructing them, and they
make sense to the student since they do explain many
observations (it is difficult to shut off air resistance and
observe a leaf and a rock falling at the same rate, thereby
allowing direct verification of the physicists view that all
objects fall at the same rate), and (c) they are resistant to
change.
Some researchers make distinctions between weak and
strong restructuring of misconceptions. For example,
young childrens eventual conception of the meaning of
alive is considered strong restructuring [36]; prior to
correctly conceptualizing alive, children believe that to be
alive means being active (people and animals are alive by
this criterion, but not lichen), or being real or seen (Carey
conveys an incident whereby her 4-year-old daughter once
proclaimed that it is funny that statues are dead even though
you can see them, but Grampas dead because we cannot
see him any more [36]). Other distinctions include intentional versus nonintentional conceptual change, where the
former is characterized by goal-directed and conscious
initiation and regulation of cognitive, metacognitive,
and motivational processes to bring about a change in
knowledge [37].
In the misconceptions view it is generally accepted that
some degree of dissatisfaction is needed for someone to
replace a misconception with a more scientifically appropriate form. As Carey argues, Without doubt, the process
of disequilibration is an important part of the process of
conceptual change [36]. Almost three decades ago Posner
and co-workers [33,38] formulated a theory of conceptual
change with four components that needed to be present
for an individual to abandon a misconception in favor
of a scientific concept: (1) dissatisfaction with a current
concept, (2) intelligibility of the new concept, (3) initial
plausibility of the new concept, and (4) usefulness of the
new concept in reasoning and making predictions about
phenomena. Strike and Posner [39] have since revised their
initial views somewhat to include issues related to the
learners conceptual ecology as well as developmental
and interactionist considerations.
3. Developing and evaluating instructional strategies to

address students misconceptions
What instructional interventions are most effective for
helping students overcome stubborn misconceptions?
Since a primary goal of science instruction is to teach
students the major concepts in each discipline as well as
how to apply those concepts to solve problems, considerable effort has gone into research to design, evaluate, and
refine curricular interventions to target students stubborn
misconceptions. This line of research builds upon the
established catalog of common misconceptions in the
research literature and uses a cyclical process to devise
instructional strategies. Most instructional techniques begin
by making students aware of their misconceptions (e.g., by
demonstrating to students inconsistencies in their own
reasoning across contextssee discussion of conceptual
change below), and then guiding students through a
series of activities or reasoning exercises to reshape their
concepts to accommodate scientific concepts (see, e.g.,
Refs. [21,2631]). Other techniques guide students
toward adopting better scientific models via teaching
interviews [32].
B. Theoretical framework
There are three main theoretical viewpoints about conceptual understanding (including conceptual development
and conceptual change) in the science education community, as summarized below.
1. Naive theories or misconceptions view
This view contends that as students gain knowledge
about the world (either through formal schooling or
informally), they build naive theories about how the
physical world works, and that often these naive theories
contain misconceptions that contradict scientific concepts
[4,5,33] (see Ref. [34] for a review). Hence, students do not
come to class as blank slates upon which instructors can
write appropriate scientific concepts [35]. For example,
children observe leaves fluttering down from tree branches
and rocks thrown from a bridge onto a stream below, and
2. Knowledge in pieces or resources view

Another view of students knowledge and conceptual
change offers a different architecture of concepts.
According to the knowledge in pieces view [4042],
020119-3
students knowledge consists of smaller grain-size pieces

that are not necessarily precompiled into larger concepts.
Students activate one to several pieces of knowledge in
response to context and reason with them on the fly. For
example, individuals of all ages [43] often state that it is hot
during summer because Earth and the Sun are in closer
proximity than during winter. This is interpreted by misconception advocates as a misconception that likely formed
from thinking about the path of Earth around the Sun as an
exaggerated ellipse. The pieces view is that perhaps the
individual generated this explanation on the spot by
searching memory for some reasonable explanation, and
coming up with the knowledge piece closer means
stronger; for example, moving closer to a fire makes
the heat one feels radiating from it stronger, and moving
closer to the stage in a rock concert makes the music louder
[44]. The pieces view is similar to Hammers resources
view [44], where resources are small units of beliefs or
thought whose correctness depends on the context in
which they are applied, or to Minstrells [45] facets.
Proponents of the pieces view argue that there are two
major difficulties with the misconceptions view that are not
present with the pieces view [40,42,44,4648]. The first is
that a misconception is thought of as a compiled cognitive
structure that an individual can employ to reason about a
situation, which implies that the individual should show
consistency in applying the misconception across similar
contexts. Interviews with students as well as assessments
reveal that slight changes in physics contexts can lead to
radically different answers, reasoning patterns, and/or
explanations [20,49,50], suggesting that novice knowledge
is highly context dependent and not stable as a misconceptions view would suggest. The second difficulty is that,
although misconceptions are constructed over time from
experiences and observations, and therefore constructivist
in nature, the misconceptions view does not provide an
account of how misconceptions eventually evolve into
correct scientific concepts. A recent publication, however,
argues that the two camps are not as far apart as some might
think [51].
ontological categories and that many of the difficulties

in student understanding are due to categorizing processes
such as heat and electric current into the matter or things
category. This would imply that the way to help students
overcome misconceptions is to have them change their
miscategorized knowledge into the appropriate ontological
categories, but misconceptions are difficult for students
to dislodge precisely because it is hard for students to
recategorize across different ontological categories. For
example, students tend to think of forces as things rather
than as the interaction between two objects, and hence talk
of forces as being used up as gasoline is used up in a car.
One well-known example of this is students discussing the
forces on a coin thrown vertically up in the airthey
mention the force of the hand as one of the forces
possessed by the coin while it is rising, a force which gets
used up when the coin reaches the top of its trajectory, after
which the force of gravity takes over to make the coin
fall [4].
Recently, this view has been challenged [21]. Gupta and
collaborators argue that both experts (in talks delivered to
other experts in their profession) and novices (when
discussing ideas with their peers) are able to traverse
between ontological categories without confusion in order
to explain complex phenomena. For example, Gupta et al.
cite physics journal articles in which physicists mix
ontological categories to build a coherent argument, such
as discussing a pulse emerging from a sample and having a
distinct peak as if it were a thing. In addition, they argue
that novices do not show stable ontological categories,
showing examples of novices speaking about electric
current as a thing and one hour after instruction using
much more sophisticated processlike descriptions of current. Finally, they also argue that crossing ontological
categories is extremely common in how we communicate
(e.g., phrases such as anger being bottled up, or having a
cold that cannot be shaken, treating emotion and sickness
as if they were things.)
3. Ontological categories view
1. Contexts
A third perspective of novice knowledge and conceptual

change is the ontological categories view. This view,
attributed to Chi and co-workers [18,19,23,52,53] and
building on previous work by Kiel [54], argues that
students naive conceptions are due to miscategorizing
knowledge and experiences into inappropriate ontological
categories, where ontological categories are loosely defined
as the sorts of things there are in the world. Examples of
ontological categories are material things like objects,
temporal things like events, and processes like the emergent
phenomenon that gives rise to such things as flocking
behavior in birds. The ontological categories view argues
that students sort knowledge into distinct and stable
Research on identifying and documenting misconceptions is done either by administering assessments designed
to probe students views in contexts where concepts need to
be applied or discussed or by conducting clinical interviews
with students about a context in which persistent conceptual errors seem to be prevalent. Clinical interviews has
been the methodology most used in probing students
conceptual architecture in memory; during clinical interviews, which typically last 1 hour, a researcher probes a
students conceptual understanding in a target topic through
a series of interviewer-led questionsquestions that often
are open ended and guided by the students responses to
previous questions. Because interview studies result in
C. Methodology (data collection or sources

and data analysis)
020119-4
large amounts of data that are difficult to analyze and

interpret, studies using this methodology rely on small
numbers of students. Studies evaluating the efficacy of
interventions for helping students overcome misconceptions are engineered (i.e., devised somewhat by trial and
error based on best guesses from research and teaching
experience) using a cyclic process where initial best
guesses are made in designing an initial intervention, then
trying the intervention and evaluating its effectiveness, and
then revising and starting the cycle again until learning
outcomes are achieved.
between the ontological categories view and the other

two in terms of explaining the nature of misconceptions and
how to help students overcome them. It is, therefore,
important to keep in mind these unresolved disagreements
in the research community as some salient findings are
presented below.
2. Participants
Misconceptions research in STEM has been conducted
with students of all ages and in various contexts such as
large undergraduate introductory courses, as well as high
school physics courses. Although not discussed in this
synthesis, misconceptions on physical science concepts
have also been studied with middle school and elementary
school children. Recently, there have been an increasing
number of studies exploring conceptual understanding
among upper-division physics students as well as graduate
students (see, e.g., Refs. [55,56]).
3. Data sources and analysis
Data sources for misconception studies come from
students performance in assessment questions or from
transcripts of clinical interviews. In cases where an intervention or teaching approach is being evaluated, assessments of misconceptions are administered to students
prior to, and following, the intervention and differences
between the postscores and prescores are analyzed. Data
from interview studies are analyzed or interpreted using
grounded theory [57], defined as developing a theory by
observing the behavior of a group (in this case, students) in
concert with the researchers insights and experiences; any
theory emerges from data as opposed to formulating
hypotheses which are then tested.
D. Findings
Discussing research findings in conceptual understanding can be a controversial issue since, as discussed above,
there are three theoretical perspectives describing the
nature of concepts in memory and conceptual change.
Misconceptions and ways of overcoming them are central
themes according to the misconceptions theoretical view,
whereas the knowledge in pieces or resources view is not in
agreement with the robustness of misconceptions or with
how one might go about overcoming them; proponents of
this view would argue that, as typically defined, students do
not possess misconceptions, but rather compile knowledge
pieces on the spot to reason about phenomena, and thus the
misconceptions that emerge for us to observe are highly
context dependent. There are also distinct differences
020119-5
1. Misconceptions
Students possess misconceptions that are deeply
rooted and difficult to dislodge [34,35]. An abundance
of misconceptions have been identified across a wide
range of physics topics in undergraduate physics (for a
good review prior to 2000, see Ref. [10]). Often,
misconceptions seemingly disappear and are replaced
with scientific concepts following instruction only to
reappear months later.
Several intervention strategies, largely based on the
conceptual change framework of Strike and Posner
[38], have proven effective at helping students overcome misconceptions. Some strategies, such as one
that has been successfully used by the University of
Washington Physics Education Research group for
many years, are based on a cyclic process that begins
by identifying misconceptions in a physics topic, then
designing interventions based on previous research,
instructor experiences, or best guesses, then piloting
the intervention and evaluating its success with posttests that measure transfer to related situations, then
refining the intervention over other cycles until
evidence is obtained that the intervention works. It
should be pointed out that even the most successful
groups at this type of work (e.g., McDermott, Herron,
and Shafer at the University of Washington) will freely
admit that devising effective instructional strategies to
combat misconceptions is often a slow, painstaking
task requiring multiple triesnot unlike engineering a
solution to a complex problem. Other approaches,
such as Clements bridging analogies [26,58,59],
start with anchoring intuitions, which are strong and
correct understandings that students possess, and
devising interventions that attempt to bridge from
students correct intuitions to related contexts in
which students display misconceptions. Yet another
very effective approach [29] uses interactive lecture
demonstrations to help students overcome prevalent
misconceptions involving Newtons third law (i.e., the
belief that heavy or fast moving objects exert larger
forces on light or stationary objects when the two
interact via collisions) by displaying in real time the
force exerted by mutually interacting carts during a
collision under different conditions (e.g., heavy cart
colliding with a light cart or moving cart colliding
with a stationary cart, etc.). After experiencing the
interactive lecture demonstrations intervention,
students retain appropriate understanding of Newtons
third law months afterwards.

Curricular interventions and assessments (both to be
discussed in different sections) have emerged based on
misconceptions research.
Classroom instruction has changed as a result of
misconceptions research, with many active learning
strategies (e.g., the use of polling clicker technologies to teach large lecture courses) being practiced
that have been shown more effective than traditional instruction at helping students overcome
misconceptions.
The pieces or resources and ontological categories
views are attempting to map human cognitive architecture, which if successful can help in designing
effective instructional strategies.
Despite the previous bullet point, designing an effective intervention to help students overcome a particular misconception can be elusive [43,60].
2. Knowledge in pieces or resources
Students possess knowledge pieces, also referred to as
phenomenological primitives (or p-prims), or resources, that are marshaled to reason about physics
contexts. These pieces of knowledge can be recalled
singly or in groups and compiled in different ways in
real time in response to different contexts. It is
common to observe that contexts that are considered
equivalent, perhaps even nearly identical by experts,
are viewed as different by students and different
knowledge pieces are brought to bear to reason about
them. Students knowledge is more dynamic in the
pieces or resources view than in the misconceptions
view. As expertise is developed with time and experience, there is refinement of the knowledge in
memory whereby knowledge pieces that repeatedly
prove effective to recall and apply to particular
contexts are compiled into scientific concepts. That
is, repeated rehearsal of knowledge pieces recalled
and compiled to deal with similar situations leads to
the formation of stable scientific concepts in experts.
3. Ontological categories
Misconceptions stem from students categorizing scientific ideas into inappropriate categories [51,53]. For
example, processes are placed in the things category
(e.g., electric current is thought of as fuel that is
used up in light bulbs).
Misconceptions are relatively easy to fix when they
involve modification within the same category, but are
difficult to fix when they involve modifications across
categories [23,61].
Instructional strategies designed to help students
overcome misconceptions by recategorizing into appropriate ontological categories have shown promise
(see, e.g., Refs. [28,6269]).
2. Limitations
Designing definitive experiments that falsify one of
the theoretical views remains elusive, hence the debate
among proponents of the three theoretical views
continues.
Although many misconceptions have been cataloged
across both introductory and advanced physics topics,
it is daunting to ever achieve a complete list.
F. Areas for future study
Research on identifying and documenting misconceptions has been progressing for several decades and has
covered an extensive range of physics topics [10], so future
research in this area is limited to alternate populations
(e.g., upper-division students, see [70]) and yet-to-beinvestigated topics. Opportunities for continued research
on the nature of student thinking and reasoning exist,
including how students ideas progress over time. There is a
pressing need for studies to help articulate general instructional strategies for guiding students to adopt scientific
conceptions, especially when those conflict with students
existing conceptions. Another promising area for future
research is to design experiments to test the three competing viewpoints outlined in the theoretical framework in
order to arrive at a more unified view.
III. PROBLEM SOLVING
E. Strengths and limitations of conceptual

understanding research
The strengths and limitations of this body of research
include the following.
1. Strengths
Misconceptions research has raised consciousness
among instructors about students learning difficulties,
and about a misconception prevalent among instructors, namely, that teaching done in a clear, elegant
manner, even charismatic instructors, quite often does
not help students overcome misconceptions.
In addition to research on conceptual understanding,

another key focus of physics education research on student
learning is problem solving, likely because problem solving
is a key component of most physics courses. As with other
studies of student learning, this research focuses first on
documenting what students do while solving problems, and
follows with the development and evaluation of instructional strategies to address student difficulties. Research
also focused on efforts to develop students abilities to
think like a physicist [7173]. Since problem solving is a
complex cognitive process, this area of research also has a
strong overlap with the section on Cognitive Psychology
020119-6
(Sec. VI). For additional reviews of research on physics

problem solving, see Refs. [74,75]. For a historical overview of early pioneers in the field of physics education
research (such as Karplus and Arons), refer to the account
by Cummings [76].
solving [100,101], as well as whether problems use

numerical versus symbolic values [102].
The research in problem solving include the following
categories and questions.
1. Expert-novice research
What approaches do students use to solve physics
problems? How are the problem-solving procedures used
by inexperienced problem solvers similar to and different
from those used by experienced solvers? How do experts
and novices judge whether problems would be solved
similarly? Early studies of physics problem solving investigated how beginning students solve physics problems and
how their approaches compare to experienced solvers, such
as professors [7783]. This area of research also includes
categorization studies to infer how physics knowledge is
structured in memory [8486].
2. Worked examples
How do students study worked-out examples? How do
students use solutions from previously solved problems
when solving new problems? How do students use instructor solutions to find mistakes in their problem solutions to
homework and exams? What features of worked-out
problem solutions facilitate student understanding of the
example? This body of research explores how students use
worked-out problem solutions or previously solved problems to solve new unfamiliar problems [87,88]. It also
includes how students use instructor solutions to selfdiagnose errors in their own problem solutions [89,90],
and how to design effective examples [91].
3. Representations
What representations do students construct during
problem solving? How are representations used by students? What is the relationship between facility with
representations and problem-solving performance? What
instructional strategies promote students use of representations? How do students frame problem-solving tasks?
This research explores the use of external representations
for describing information during problem solving, such as
pictures, physics-specific descriptions (e.g., free-body diagrams, field line diagrams, or energy bar charts), concept
maps, graphs, and equations. Some studies focus on what
representations are constructed during problem solving and
the manner in which they are used [9296], whereas other
studies explore the facility with which students or experts
can translate across multiple representations [9799]. How
students frame the problem-solving task impacts problem
4. Mathematics in physics
How are the mathematical skills used in physics courses
different from the mathematical skills taught in math
courses? How do students interpret and use symbols
during quantitative problem solving? This area of research
explores how quantitative tools from mathematics courses
are applied during physics problem solving. Some examples include the use of symbols in equations to represent
physical quantities [103106], vector addition [107], arithmetic, algebra, geometry, calculus (e.g., integration) [108],
and proportional reasoning [109].
5. Evaluating the effectiveness of instructional
strategies for teaching problem solving
How does instructional strategy X [e.g., cooperative
group problem solving] affect students problem solving
skills? To what extent do conceptual approaches to
problem solving influence students conceptual understanding of physics? How do students interact with online
computer tutor systems? What features of Web-based
homework systems successfully enhance students problem
solving skills? Several instructional strategies have been
developed and tested, including the use of alternate types
of problems [110114], adopting an explicit problemsolving framework (i.e., a consistent sequence of problem
solving steps) [115], conceptual approaches [116118],
cooperative group problem solving [119], and computer
homework or tutor systems to help students become better
problem solvers [71,120,121].
B. Theoretical frameworks
This section identifies a few prominent theories about
learning and problem solving from cognitive science and
educational psychology. Although these frameworks are
viewed as useful and relevant for PER, problem-solving
researchers in PER often do not clearly define or draw upon
a theoretical basis for their research studies. The frameworks reviewed here are intended to provide a starting point
for discussion. The following frameworks are included:
information-processing models, problem solving by analogy, resources model, and situated cognition.
1. Information-processing models
According to one theory of human problem solving
[122124], problem solving is an iterative process of
representation and search for a solution. This theory defines
a mental state called the problem space consisting of a
persons available knowledge and their internal representation or understanding of the task environment, including
the initial state (given information), goal state or target,
and appropriate operations that can be employed. After
020119-7
representing the problem, a solver engages in a search

process during which they select a goal and a method to
apply (such as a general heuristic), applies it, evaluates the
result of this choice, modifies the goal or selects a subgoal,
and proceeds in this fashion until a satisfactory solution
is achieved or the problem is abandoned. Gick [125]
expanded upon this model to include the role of schema
activation. If constructing an internal representation of the
problem activates an existing schema or memory framework for solving the problem, a solution is implemented
immediately. If no schema is activated, then the solver
engages in general search strategies [125].
Within this theory of problem solving, Newell and
Simon [123] also describe the organizational structure of
memory as an information-processing system consisting of
two branches: short-term or working memory (STM) and
long-term memory. Short-term memory is constrained and
can only hold a small amount of information for a limited
time, whereas long-term memory is essentially unlimited
[124]. However, in order to access information stored in
long-term memory, it must be activated and brought into
working memory. If problem information and activated
knowledge exceed the limits of STM, a solver may
experience cognitive overload [126], which interferes with
attempts to reach a solution. To alleviate this effect,
problem information is often stored externally (e.g., written
down on paper) or processed with the aid of a tool (e.g.,
computer) in order to free up space in working memory
to devote to the task. As a problem solver becomes more
proficient, knowledge and procedures may become
chunked and some skills become automatic, which also
optimizes the capacity of STM.
emphasize the role of specific example problems for

guiding the application of abstract principles [131,132].
Some studies have explored the value of using isomorphic
problems in helping students draw analogies between
problems [134]. For a more comprehensive review of these
views, see Reeves and Weisberg [133].
2. Problem solving by analogy

Instructors report that students often look for similar
example problems in physics textbooks when attempting to
solve homework problems. This process of using a similar,
familiar problem to solve a new problem is referred to as
analogy. There are three main views of analogical transfer,
including structure mapping [127,128], pragmatic schema
view [129,130], and exemplar view [131,132]. These views
agree that there are three main criteria for making use of a
known problem: (1) a person must be reminded of the
previous problem and sufficiently remember its content,
(2) they must compare and adapt the attributes from one
problem to the other (what Gentner calls structure mapping), and (3) the solver must evaluate the effectiveness of
this transfer for solving the new problem [133]. To be
useful for future problem solving, a person must also
abstract common features and generalize from the examples. The work of Holyoak and Koh [130] suggests that
similarity can be both at a surface level and at a structural
level, and while surface similarity can aid with reminding,
it is structural similarity that is important for appropriately solving the new problem. Exemplar-based models
3. Resources model and epistemic games

Research on how students use and understand mathematical symbols and equations in physics has built upon a
theoretical framework called the resources model [44,135].
This model describes knowledge in terms of students
resources for learning, both conceptual and epistemological in nature [44]. This model helps to explain why
students might have the requisite knowledge and skills for
solving a problem but fail to activate it in a particular
context, and suggests that instruction should be designed to
make productive use of students resources. Tuminaro
[136] and Tuminaro and Redish [106] extended this
resources model to give a detailed framework for analyzing
students application of mathematics to physics problems.
They identified six epistemic games or problem-solving
approaches that students participate in while using math in
physics, including mapping meaning to mathematics,
mapping mathematics to meaning, physical mechanism,
pictorial analysis, recursive plug and chug, and transliteration to mathematics.
4. Situated cognition
Some research on instructional strategies for teaching
problem solving (e.g., cooperative problem solving) is
based on the situated cognition model of learning. In this
view, knowledge is situated or influenced by the particular task and context in which learning is taking place
[137]. They propose that school activities should incorporate authentic tasks, and they propose a teaching method
called cognitive apprenticeship to facilitate an enculturation
process. Cognitive apprenticeship includes three main
phases: modeling (demonstrating how to do something),
coaching (providing opportunities for guided practice and
feedback), and fading (gradually removing scaffolding).
and data analysis)
1. Contexts
Basic research to understand how students solve
problems and identify common student difficulties is
typically done outside the classroom, in an experimental
or clinical setting with paid volunteers. Studies of pedagogy and curricular interventions are conducted in conjunction with a course where the data collected includes
student responses to course materials such as homework
and exams.
020119-8
2. Participants
Studies of physics problem solving frequently involve
undergraduate students in large-scale introductory courses
(algebra based or calculus based) at a single institution over
one or two semesters, sometimes longer. Studies conducted
with high school students or disciplinary majors are less
common. The experts in expert-novice studies are often
physics faculty or graduate teaching assistants (TAs) who
are experienced with teaching introductory courses.
One common data source for problem-solving research
is students written solutions to physics problems that have
been designed by the researcher(s) and/or adapted from
existing problem sources. In most cases the problems are
free response, but occasionally they are in a multiple-choice
format. These data are typically analyzed by scoring the
solutions according to particular criteria, as determined by
comparing the written solutions to features of an ideal
instructor solution. Sometimes rubrics are used to make
scoring criteria objective and improve agreement among
multiple scorers. Students written solutions are used in
curricular intervention studies to compare performance in
reformed courses to traditional courses on a set of common
assessment items.
Another widespread data source, especially for cognitive
studies, is think-aloud problem-solving interviews. To
collect such data participants are asked to solve physics
problems and verbalize their thoughts during the process
while being video- and/or audiotaped. Transcripts from
these interviews are analyzed using standard qualitative
methods such as case studies or grounded theory [57] to
elucidate common themes or problem-solving approaches
from the statements.
Studies comparing expert and novice problem solvers
have traditionally used categorization tasks that require
participants to group problems based on similarity of
solution. Early studies used card-sorting tasks where
subjects physically placed problem statement cards into
piles and the resulting category was assigned a name
[84,138]. More recent studies have used alternate categorization tasks, such as selecting which of two problems
would be solved most like a model problem [85], multiplechoice categorization formats [117], or ranking the
similarity of two problems on a numerical scale [139].
Some categorization tasks are analyzed using qualitative
methods (to elicit commonalities across card groupings),
whereas others compare quantitative ranking or the frequency of similarity judgment responses for different
groups of subjects. New approaches are emerging for
interpreting card-sorting categorization data [140].
Eye-tracking technology is a rare data source, but in
some cases it is used to study gaze patterns or fixation time
upon particular features of problems or problem solutions
presented on a screen [141143]. These problem features
could include representations (such as pictures or physics

diagrams), text, equations, and mathematical steps in
example solutions.
D. Findings
1. Expert-novice studies
Early studies of problem solving identified differences
between beginning problem solvers and experienced problem solvers in both the way they organize their knowledge
of physics and how they approach problems.
Categorization. Expert-novice studies found that
when asked to group problems based on similarity
of their solution, beginning problem solvers used
literal objects from the surface attributes of the
problems as category criteria (such as spring problems or incline plane problems), whereas experienced problem solvers considered the physics concept
or principle used to solve the problem when deciding
on problem categories, such as grouping conservation
of energy problems together [84,144]. It was later
shown that expert novice behavior is a continuum not a
dichotomy; sometimes beginning students exhibit
expertlike behaviors and experienced solvers behave
like novices [85]. Further, studies have also demonstrated that novices also rely on terminology and
variable names in the problems to make categorization
decisions, not just objects in the problem [85,145].
More recently a study examined the type of problems
needed to accurately distinguish between experts and
novices [146].
Problem-solving approaches. Think-aloud problemsolving interviews found that expert problem solvers
typically begin by describing problem information
qualitatively and using that information to decide on
a solution strategy before writing down equations
[7783]. A successful solvers strategy usually includes the appropriate physics concept or principle
and a plan for applying the principle to the particular
conditions in the stated problem [147]. In contrast,
beginning physics students typically started by writing
down equations that match given or desired quantities
in the problem statement and performing mathematical
manipulations to get an answer [84,148]. A study
showing how little good students retain from an
introductory course they took in the freshman year
when tested on the knowledge in their senior year is
sobering [149].
Metacognition. In addition, these studies concluded
that experts monitor their progress while solving
problems and evaluate the reasonableness of the
answer, whereas beginning students frequently get
stuck and lack strategies to progress further
[81,87,150152]. Explicit instruction on strategies
for evaluating a solution, such as the use of limiting
and extreme cases, has been shown to improve
performance [151].
020119-9
2. Worked examples
A common practice in problem-solving instruction is to
provide students with worked-out example problems that
demonstrate solution steps. This section briefly reviews
research on how students study example problem solutions
in textbooks, how they refer to previous examples during
problem solving, and how they use instructor solutions to
detect and correct errors in their own solutions. This
research also addresses how to design example solutions
to optimize their usefulness.
Using examples. Research conducted by Chi et al.
[87] and Ferguson-Hessler and de Jong [153] on how
students study worked-out example solutions in textbooks concluded that good and poor students used
worked examples differently when solving a new
problem. In Chi et al. [87], the labels good and
poor were determined by a students success at
solving problems after studying the worked examples
(12 isomorphic or similar problems and 7 unrelated
problems from a textbook). Good students (who
scored an average of 82% on problems) referred to
a specific line or line(s) in the example to check
procedural aspects of their solution, whereas poor
students (who only scored an average of 46% on
problems) reread the example from the beginning to
search for declarative information and a solution
procedure they could copy [87]. Ferguson-Hessler
and de Jong [153] confirmed these findings and
described the actions of good students as deep
processing of the examples whereas poor students
engaged in superficial processing.
A study by Smith, Mestre, and Ross [143] used eye
tracking to investigate what aspects of a problem solution
students look at while studying worked-out examples. They
found that although students spent a large fraction of time
reading conceptual, textual information in the solution,
their ability to recall this information later was poor. The
students eye-gaze patterns also indicated they frequently
jumped between the text and mathematics in an attempt to
integrate these two sources of information.
Self-diagnosis. When students receive feedback on
their performance on homework and exams, instructors generally expect students to reflect on and learn
from their mistakes. However, research indicates that
such self-diagnosis of errors is difficult for students. In
a study by Cohen et al. [89], only one-third to one-half
of students were able to recognize when they used an
inappropriate physics principle for a problem. These
results were affected by the support provided to the
students, indicating that having access to a correct
solution and self-diagnosis rubric helped students
explain the nature of their mistakes better than an
instructor solution alone or only access to the textbook
and class notes [90]. A classroom intervention by
Henderson and Harper [154] found that requiring
students to correct mistakes on assessments improved

student learning of content and also improved skills
relevant for problem solving, such as reflection.
Example structure and the worked-example effect.
Ward and Sweller [91] conducted classroom experiments using topics of geometric optics and kinematics
to investigate how the design of example solutions
influences learning and problem solving. They found
that under certain conditions reviewing correct example solutions was more effective for learning than
actually solving problems, what is referred to as the
worked-example effect. To optimize their usefulness,
they claim that examples must direct students attention to important problem features and minimize
cognitive load by keeping information sources separated, such as diagrams, text, and equations.
3. Representations
The term representation has multiple interpretations, but
for the problem-solving research reviewed here it is used to
refer only to concrete, external descriptions used by a
solver. Some examples include pictures or sketches, physicsspecific descriptions (e.g., free-body diagrams, field line
diagrams, ray diagrams, or energy bar charts), concept
maps, graphs, and equations or symbolic notation. Some
researchers go on to make a distinction between general and
physics-specific representations. Reif and Heller [151]
suggest that a basic description includes initial steps taken
to understand a problem, such as introducing symbolic
notation and summarizing problem information verbally or
pictorially. They separate this from a theoretical description
that specifies systems and interactions for objects using
physics concepts, such as describing motion with position,
velocity, and acceleration or describing interactions by
force vectors.
Representational format affects performance. Multiple
studies have determined that some students give
inconsistent answers to the same problem-solving
question when it is presented in different representational formats [98,99]. The four formats tested
included verbal, diagrammatic (or pictorial), mathematical or symbolic, and graphical. The pattern of
performance differed by the task and problem topic, but
performance on mathematical quizzes was usually
worse than other formats despite students preference
for calculation questions [98].
Physics-specific descriptions (might) facilitate problem solving. A clinical study by Heller and Reif [92]
found that subjects who received training materials on
procedures for generating physics-specific, theoretical
descriptions in mechanics performed better on a set of
three problems than an unguided group. The group
scored significantly higher on all four aspects of their
solutions: describing the motion of objects, drawing a
force diagram, correct equations, and the final answer.
020119-10
They concluded that these descriptions markedly

facilitate the subsequent construction of correct problem solutions (p. 206). In a study by Larkin [93], five
students who were trained to use physical representations for direct-current circuits solved significantly
more problems on average than a group of five
students who received training on generating and
combining equations to obtain an answer. The subjects
were presented with three problems, and a problem
was considered solved if the student produced a
correct solution within a 20-minute time limit. Other
studies have produced mixed results regarding the
relationship between representations and problemsolving success. Studies by Rosengrant, Van Heuvelen,
and Etkina [95,142] found that students who drew
correct free-body diagrams were much more likely to
solve a problem correctly than those who drew an
incorrect diagram or none at all, but those who drew
an incorrect diagram performed worse than students
who drew no diagram. In a study by Van Heuvelen and
Zou [96], students reported multiple work-energy
representations were useful tools for solving problems, but the researchers observed little use of the
representations on exams and could not determine
their relationship to problem-solving performance.
Similarly, Rosengrant and Mzoughi [155] found
favorable student reactions to the use of impulsemomentum diagrams during class, but no differences
in performance. De Leone and Gire [156] did not find
a significant correlation between the number of
nonmathematical representations used by a problem
solver and the number of correct answers, but the
nature of the errors made was different for representation users and nonusers.
Successful problem solvers use representations differently than unsuccessful solvers. Some studies indicate
that successful problem solvers use qualitative representations to guide the construction of quantitative
equations [157] and they use physics diagrams as a
tool to check the consistency of their work [158]. In
contrast, unsuccessful solvers either do not generate
representations or do not make productive use of them
for problem solving. Kohl and Finkelstein [159]
suggest that both novices and experts generate multiple representations during problem solving and sometimes do so in a different order, but experts can more
flexibly move between those multiple representations.
4. Mathematics in physics
Mathematics and equations are often referred to as the
language of physics. This area of study identifies
common student difficulties associated with math skills
required for physics problems and describes how equations
are used differently in the context of physics compared
to math.
Vectors. Studies have documented several student

difficulties associated with using vectors, particularly
for velocity, acceleration, forces, and electric fields
(see, e.g., Refs. [107,160162]). In particular, Nguyen
and Meltzer [107] identified widespread difficulties
with two-dimensional vector addition and interpreting
the magnitude and direction of vectors in graphical
problems.
Algebraic or proportional reasoning. Some students
have difficulty translating a sentence into a mathematical expression, and in doing so they often place
quantities on the wrong side of an equals sign. Cohen
and Kanim [109] studied occurrences of the algebra
reversal error in both algebra and physics contexts,
concluding that sentence structure was the most
influential factor in guiding translation of sentences
expressing relationships among variables into equations. More recently Christianson, Mestre, and Luke
[163] found that the reversal error essentially disappears as students practice translating algebraic statements into equations, despite never being given
feedback on whether or not their equations were
correct.
Symbols and equations. In contrast to mathematical
expressions, equations used in physics have conceptual meaning associated with symbols and the relationships among physical quantities [106,136].
Students generally have poorer performance on symbolic questions than on numeric questions, and
Torigoe [105] attributes this difference to students
confusion about the meaning of symbols in equations
and difficulty keeping track of multiple quantities.
Sherin [103,104] identified 21 different symbolic
forms for equations in physics and how students
interpret them, concluding that it is possible for
students to have mathematically correct solutions
but inappropriate conceptual understanding for the
equations.
Use of mathematical skills is context dependent. In
many cases, students have the mathematical resources
to solve problems but do not attempt to apply them in
the context of a physics class or vice versa [164166].
This suggests that the availability of resources
depends on the context in which a skill was learned
and the solvers perception of appropriate knowledge
to activate in a particular situation, referred to as
epistemological framing [136,167]. Cui, Rebello, and
Bennett [108] found that students needed prompting
and guidance to make the connection between their
knowledge of calculus and a physics problem. Sayre
and Wittman [168] found that students had a more
solid understanding of Cartesian coordinate systems
and would persist in applying this system even
when a different (polar) coordinate system was
more appropriate. They used Resource Theory and
Process/Object Theory to explain the development of
020119-11
unfamiliar ideas, suggesting a plasticity continuum

measure in which resources can range from being
plastic (uncertain about how and when to apply them)
to being solid (familiar and confident in their understanding of how to apply them).
5. Instructional Strategies
Alternate problem types. Standard problems presented
in textbooks are often well defined and arguably do
not reflect the nature of scientific thinking [74].
Several types of problems have been developed as
alternatives to these standard problems that emphasize
conceptual reasoning for realistic contexts. Some
examples include context-rich problems [110,169],
experiment problems [113,170], jeopardy problems
[114], problem posing [111], ranking tasks [112,171],
real-world problems [172], thinking problems [173],
and synthesis problems [174].
The effects of implementing alternative problems are
sometimes difficult to separate from the instructional
strategies used in conjunction with the problems. For
example, context-rich problems [110,119] result in
improved problem-solving performance when used in
conjunction with a problem-solving strategy and cooperative group problem solving. Specifically, students
made fewer conceptual mistakes, generated more useful
descriptions, and wrote equations that were consistent with
their descriptions.
Experiment problems are typically utilized during laboratory sessions, where students can explore the question
using objects similar to those in the problem [113]. The
article only describes the problems, and does not provide
research results on their effectiveness. The use of jeopardy
problems resulted in high scores on the mechanics baseline
test and the Force Concept Inventory (FCI) [114], but they
caution that other instructional curricula were also used in
conjunction with these problems.
In problem posing [111], students must generate a
problem statement that meets certain criteria (for example,
they are shown a diagram of two blocks connected by a
string and they must pose a problem that can be solved by
Newtons second law). This was found to be an effective
tool for assessing a students knowledge of physics concepts and their problem-solving skills, especially their
ability to link appropriate problem contexts with physics
principles.
Problem-solving frameworks. Explicitly modeling an
organized set of problem-solving steps and reinforcing
this framework throughout a course results in higher
course performance [92,115,175,176]. Often such
frameworks are based on the work of the mathematician Polya [177]: understand the problem, plan the
solution, execute the plan, and look back [73,178,179]
but are expanded to include explicit guidelines for
constructing an initial description of the problem.
Qualitative approaches. Some curricula emphasize

performing a qualitative analysis of a problem before
writing down quantitative equations. Some examples
include Overview, Case Study Physics [73,118,180],
hierarchical problem analyses based on principles
[181,182], and strategy writing, where a strategy
consists of the principle, justification, and procedure
needed for solving a problem [117].
Problem-solving instruction. Some studies describe
how courses can be restructured to include explicit
instruction on problem-solving skills, such as through
the use of cooperative group problem solving [119] or
a curriculum called Active Learning Problem Sheets
[183,184]. For additional instructional strategies, see a
meta-analysis presented in Ref. [185].
Computer tutors. Early uses of computers as problemsolving tutors focused on prompting students to use a
systematic approach for solving Newtons laws and
kinematics problems [186188]. These tutoring systems have expanded substantially to include sophisticated hints, guidance, and feedback [71,189,190].
Other computer-assisted programs focus on Webbased homework such as the computer-assisted personalized assignment system (CAPA) [120,191193],
Mastering Physics, and CyberTutor [121].
Studies comparing paper-and-pencil homework to
Web-based homework have found mixed results, with
some studies citing no differences in performance [194]
and others citing improved performance with Web-based
tutors or homework systems [195198]. Additional
research results about homework are summarized in Sec. V.
E. Strengths and limitations of
problem-solving research
1. Strengths
A strength of problem-solving research is that it is an
established subfield of PER with a strong history and
associations to other fields, such as mathematics and
cognitive science. As a result, we have a great deal of
information about how students solve problems and
instructional strategies to address common difficulties. In
addition, we can make use of established research methodologies from cognitive science and psychology when
conducting research in physics education.
2. Limitations
Many of the conclusions from problem-solving research
have been drawn from studies with a low number of
subjects, a limited range of problem types and topics,
and inconsistent measures of problem-solving performance. In addition, the complex nature of problem
solving and variability in problem features make it
difficult to isolate particular factors that may be responsible
020119-12
for improvements (or decrements) in problem-solving

performance.
Low numbers of subjects. Some early studies had very
few subjects and/or the results have not been replicable on a larger scale (see, for example, Ref. [138] on
revisiting categorization). Also, few studies have
investigated problem-solving approaches of transition
stages in the expert-novice continuum, such as undergraduate majors and graduate students [199,200].
Limited range of problem-solving tasks. In many
expert-novice studies, the experts are solving exercises that are simple for them, not novel problems they
have not seen before. Very few studies have investigated the approaches that experts take on complex
problems (see, for example, Ref. [152]). There is little
research comparing problem solving across physics
topics, for example, contrasting problem solving in
mechanics with problem solving in electricity and
magnetism or solving problems in upper-division
physics courses.
Inconsistent problem-solving measures. There is no
widespread, consistent measure of problem-solving
performance [201]. In most of the studies reviewed
here, problem solving was measured using vague
techniques such as mean number of correct answers,
exam scores, or course grades without presenting
detailed information about the way in which problem
solutions were scored. There is some current research
to develop a valid and reliable measure of problem
solving, but for now this is still a limitation of many
studies [202204].
Failure to systematically consider problem features.
Problem solving is a complex topic of study, and most
research studies do not systematically explore the
effects of changing individual problem or solution
features on problem-solving performance. For example, problems can differ along the following dimensions: physics concepts and principles required to
solve it (or a combination of multiple principles), the
format of the problem statement (text, picture, diagram, graph), the mathematics required for a solution,
values provided for quantities (numeric) or absent
(symbolic), presence of additional distracting information, context (e.g., real objects like cars or superficial objects like blocks), and the familiarity of the
problem context (e.g., sports compared to nuclear
particles).
most prominent gaps in research on problem solving

include research on worked examples, multiple representations, reducing memory load, and adoption of reformed
instruction. In particular, the research on the workedexample effect in physics is sparse, and there are few
guidelines for how to best design instructor solutions.
Research on multiple representations is both sparse and
contradictory, with little evidence regarding the relationship
between use of representations and problem-solving performance. Another area that would benefit from future
study is developing strategies for effectively reducing
memory load while still highlighting important aspects
of problem solving. Although there are several instructional
strategies and curricula for teaching problem solving,
adoption of these practices is not particularly widespread
and this warrants additional study.

This review identified five key areas of existing research
on problem solving: expert-novice research, worked examples, representations, the use of mathematics in physics,
and evaluating the effectiveness of instructional strategies
for teaching problem solving. These areas include many
opportunities for continued research; however, some of the
IV. CURRICULUM AND INSTRUCTION

IN PHYSICS
An abundance of instructional methods and curricular
materials have been developed to span all aspects of the
standard physics course: lecture, recitation, and laboratories. Some of these instructional reforms aim to incorporate
active engagement of students in traditional courses,
whereas other reforms involve comprehensive structural
changes, such as combining lecture, recitation, and labs
into a single class environment. Advances in technology
have also introduced classroom polling technologies into
lectures, computers and sensors for collecting and analyzing laboratory data, Web-based homework and tutoring
systems, and computer animations or simulations of physical phenomena. As a result, the development and evaluation of instructional innovations continues to be an active
area of PER.
Several PER-based curricular materials have been
developed and tested for their effectiveness. There are
also several text resources available for physics instructors to learn more about PER-based teaching strategies.
Redishs [205] book Teaching Physics with the Physics
Suite gives an overview of several research-based instructional strategies, curricula, and assessments, and also
includes a chapter on cognitive science. Books that
provide hints for teaching specific physics topics include
Aronss book A Guide to Introductory Physics Teaching
[206] and Knights Five Easy Lessons: Strategies for
Successful Physics Teaching [207]. For a comprehensive
review of research on active-learning instruction in
physics, see Ref. [208].
All examples of instructional reform presented here are
(either implicitly or explicitly) related to the overarching
research question of evaluating the effectiveness of instructional strategies and materials for teaching physics, and
020119-13
determining the conditions under which the instruction

does or does not work. We will, therefore, forego listing
specific research questions below since nearly all would be
of the Did it work, and under what conditions? type.
Strategies include lecture-based instruction, recitations, and
laboratory-based strategies; changes to the overall structure
of a class; and general curricular materials such as textbooks and simulations. Many of these examples have been
termed interactive engagement methods in comparison to
traditional, passive lecture methods [209].
1. Lecture-based methods
These instructional methods seek to enhance the lecture
format of a physics course by making in-class experiences
interactive through student-student interactions and
student-instructor interactions [210]. Reforms to lecture
often include opportunities for students to discuss topics
with their classmates, like in active learning classes supported by classroom polling technologies [211213], and
Interactive Lecture Demonstrations [214]. Another instructional method is to require students to complete prelecture
assignments to better prepare them for lecture, such as
answering conceptual questions in Just-in-Time Teaching
[215] or viewing a presentation such as multimedia learning
modules [216218]. Other lecture-based approaches promote self-reflection about learning, such as requiring
students to keep a journal and/or submit weekly reports
describing what they have learned and what they are still
confused about [219221].
4. Structural changes to classroom environment

Some methods of instructional reform involve alterations
to the standard classroom structure. One example is
problem-based learning, which modifies the lecture setting
to be appropriate for cooperative group work [236,237].
Several other examples revise the course to integrate some
aspects of lecture, labs, and recitations into a single setting
as in Physics by Inquiry [238,239], Workshop Physics
[240,241], Studio Physics [242246], Student-Centered
Active Learning Environment for Undergraduate
Programs (SCALE-UP) [247], and Technology-Enabled
Active Learning (TEAL) [248250].
5. General instructional strategies and materials
This class of reform includes physics textbooks that have
been written to incorporate results from physics education
research such as Understanding Physics [251] and Matter
and Interactions [252] and books about teaching physics
such as Redishs [205] Teaching Physics with the
Physics Suite and Aronss [206] A Guide to Introductory
Physics Teaching. It also includes computer-based curricula
such as animations and interactive simulations of physical
phenomena [253255] and instruction that incorporates
computer programming experiences into the course
[256,257]. Other curricula cited include high school materials such as Modeling Instruction [258] and Minds on Physics
[259], since these are sometimes adopted for undergraduate
courses for nonscience majors [260].
2. Recitation or discussion methods

These approaches seek to make traditional, passive
recitation sessions more interactive and collaborative.
Curricular materials include the Tutorials in Introductory
Physics developed at the University of Washington [27] and
adaptations thereof [222,223], such as Activity-Based
Tutorials and Open-Source Tutorials at the University of
Maryland [224226]. Another recitation method utilizes
collaborative learning in discussion sessions, such as
Cooperative Group Problem Solving at the University of
Minnesota [110,119].
3. Laboratory methods
The introduction of technological tools for collecting
and analyzing data has transformed the physics laboratory and spawned several curricula to accompany these
changes, including RealTime Physics [227230] and
video analysis software [231]. Other laboratory materials
focus on developing scientific abilities (e.g., designing
experiments, testing hypotheses) such as Investigative
Science Learning Environments [232,233], Computerbased Problem-Solving Labs [234], and Scientific
Community Labs [235].
Most modern theories of instructional design for science

courses are based on constructivism and its associated
theories, such as situated, sociocultural, ecological, everyday, and distributed cognition [261]. The constructivist
view of teaching and learning emphasizes the active role
that learners take by interacting with the environment and
interpreting information in relation to their prior understanding and experiences [262,263]. Constructivism has
roots in early theories of education; for example, the notion
that experience and prior knowledge influence learning was
discussed by the education theorist Dewey [264] and
apparent in Piagets theories of cognitive equilibration
via assimilation and accommodation. For a historical
review of constructivist theories including a discussion
of Piaget, see Ref. [263].
Instructional strategies and materials for teaching undergraduate physics have incorporated constructivist theories
by reforming traditional styles of teaching (passive, lecturebased methods) to be more student centered and active,
approaches collectively referred to as interactive engagement methods [209]. Examples of instructional strategies
and materials to make lectures more interactive are
described in the Sec. IV D 1.
020119-14
Design principles for creating constructivist learning

environments include assigning open-ended tasks (e.g.,
problems) that are authentic and challenging, giving
students opportunities to work collaboratively with their
peers, and providing appropriate scaffolding for activities
[265]. Physics curricula that incorporate alternate problems
and cooperative group work include problem-based learning [236] and cooperative group problem solving
[110,119]. Other instructional strategies that also utilize
teamwork and cooperative learning in the classroom
include Physics by Inquiry [238,239], Tutorials in
Introductory Physics [27], Workshop Physics [240,241],
Studio Physics [242,243,245,246], SCALE-UP [247], and
TEAL [248,249].
Another constructivist-based approach to instruction is
the use of learning cycles [266,267]. A three-phase
learning cycle developed for the Science Curriculum
Improvement Study (SCIS) included exploration, concept
introduction, and concept application [266]. An extension
of this work is the 5E cycle of engagement, exploration,
explanation, elaboration, and evaluation [262]. Both
approaches include an initial exploratory period in
which students participate in hands-on activities before
the formal introduction of concepts. Several physics
laboratory curricula incorporate this aspect of early
exploration, such as Physics by Inquiry [238,239] and
Investigative Science Learning Environments [232]. The
upper-division curricula Paradigms in Physics also cites
the use of learning cycles [268].
research methodologies. The development of curricula

is more descriptive and often begins with identifying
common student difficulties or misconceptions, so it lends
itself to qualitative methods such as interviews with
students or faculty. Examples include the process undergone to develop the Physics by Inquiry and Tutorials in
Introductory Physics at the University of Washington
[5,162,269271] or the procedures to validate clicker
questions at The Ohio State University [211].
Evaluating the effectiveness of reforms made to large
introductory physics courses lends itself to quantitative,
statistical methods and the use of quantitative measures.
Common data sources for evaluating the effectiveness of
instructional strategies or materials include student
responses to standard course assignments, such as performance on course exams, laboratories, and homework.
Sometimes rubrics are written to facilitate scoring for
open-ended items [272]. For instructional reforms that
have taken place within the past two decades, a common
data source is students performance on pre-post concept
inventories such as the Force Concept Inventory and Force
and Motion Concept Evaluation (FMCE) or attitudinal
surveys. (For a comprehensive discussion of inventories,
see Sec. V.) Quantitative data analysis methods are utilized
to compare scores, and a commonly cited pretestpost-test
measure is the normalized gain [209]. Occasionally qualitative research methods are utilized to examine student
discourse and student interactions during a reformed class
with the aid of video and audio recording [249].

and data analysis)
D. Findings
1. Contexts
1. Lecture-based methods
Studies of the effectiveness of particular instructional

strategies and/or materials are typically conducted in the
context of physics courses. Oftentimes students performance in a reformed course is compared to other sections of
a similar course taught traditionally, or to past years in
which the course was taught differently. Comparisons are
typically made between courses within a single institution,
but occasionally researchers make cross-institutional comparisons [209].
2. Participants
Studies of reformed instruction frequently involve
(algebra based or calculus based), and in very few instances
there are reforms made to upper-division undergraduate
physics courses or graduate student courses. Refinements
to instructional methods can take place over a period of
several years with multiple cohorts of students.
The development and evaluation of reformed curricula
typically involves a mixture of quantitative and qualitative
Peer discussions and classroom communication

systems. Several methods have been developed to
encourage active student involvement during a lecture
class, including active learning via classroom pollingfacilitated instruction [212] and Peer Instruction
[213]. In these approaches, students engage in a
questioning cycle that typically includes viewing a
question presented by the instructor, discussing ideas
with classmates, responding to the question, and
engaging in a classwide discussion. In questionnaires
and interviews, students reported having a positive
attitude toward active learning using polling technologies and felt that they learned more than they would
have in a passive lecture class [212]. Studies of the
effectiveness of Peer Instruction cite increased conceptual understanding and problem solving performance [273] and decreased course attrition [274].
The nature and quality of the questions asked is also
important [211,275]. Qualitative questions and questions
that require reasoning promote discussion more than
questions requiring only a calculation or recall of information [275], and students and instructors both report a
preference for such conceptual questions over numerical
020119-15
problems [212,276]. Beatty et al. [275] suggest several

ways in which questions can be crafted to include subtle or
common student conceptions, such as sequencing two
questions to look similar but require different reasoning
in an oops-go-back format [275]. Researchers at The
Ohio State University went through a process of validating
several clicker questions with a sequence of student interviews, faculty interviews, and classroom response data
[211]. They identified some situations in which students
perspectives were not anticipated by the question designers,
highlighting the importance of validating course materials.
The manner in which credit is assigned for responses has
been shown to affect the nature of students conversations
[277]. In high-stakes classrooms, group conversations are
dominated by high-ability students, whereas low-stakes
classrooms promote discussion among students of varying
ability levels.
These lecture-based methods frequently employ student
response systems such as clickers or other polling or
communication devices (see, for example, Ref. [278]).
Studies of electronic response systems for gathering
student answers indicate that it is not the technology that
is important, but the way in which it is used [276,279,280].
Clickers offer advantages for instructors, such as acquiring
an immediate gauge of student understanding and keeping
an electronic record of response patterns [280]. For a
review of research on clickers and best practices for their
use, see Ref. [279].
Interactive lecture demonstrations (ILDs). This curriculum follows a sequence in which students make a
prediction about what they expect to see in a physical
experiment or demonstration, discuss their prediction
with peers, observe the event, and compare the
observation to their prediction [214]. Each stage of
the ILD is guided by questions printed on worksheets.
Studies evaluating the effectiveness of ILDs indicate
that they significantly improve learning of basic
physics concepts measured by the FMCE [29] and
that the prediction phase of classroom demonstrations
is particularly important for forming scientific concepts [281].
Prelecture assignments. One prelecture method,
called Just-in-Time teaching (JiTT), requires students
to submit answers to conceptual questions electronically before class [215]. The instructor uses these
responses to gauge student understanding and incorporates them into the lecture. JiTT has been shown to
improve students understanding of concepts such as
Newtons third law [282]. In another prelecture
method, students view multimedia learning module
(MLM) presentations with interspersed questions
[283]. Students who viewed MLMs performed better
on assessments of basic content administered both
immediately after viewing the MLMs as well as two
weeks later compared to students who read the same
material in a static text format [217,218], and MLM

viewers also performed better on before-lecture
assessment questions indicating the presentations
improved students preparation for class [216]. In
addition, when MLMs were implemented, students
perceived the course as less difficult, they had a more
positive attitude toward physics, and they found
lectures more valuable since the MLMs allowed class
time to be spent on activities to refine their understanding rather than covering basic content [217].
Similar results were found in a different implementation of the MLM approach [284,285].
Reflection on learning. One practice to facilitate
students reflection on their learning is requiring
them to keep a journal and/or submit weekly reports
[219221]. In those studies, students were asked to
write down every week what they learned in physics
class, how they learned it, and what they still had
questions about. Students received credit for submitting responses that were similar in weight to homework. When students weekly journal statements were
coded across several aspects evaluated for their
appropriateness (see Ref. [221] for the code criteria),
the researchers found a relationship between performance and reflection quality, where students with high
conceptual gains on concept inventories tended to
show reflection on learning that is more articulate
and epistemologically sophisticated than students with
lower conceptual gains [221].
2. Recitation or discussion methods
In physics courses, a recitation or discussion session
refers to a smaller classroom environment that meets once
or twice a week, typically consisting of 1530 students and
taught by a graduate or undergraduate teaching assistant. It
is intended to provide students with individualized help and
feedback, such as with conceptual activities or problemsolving practice. In traditional recitations the TA would
solve homework problems on the board, whereas reformed
sessions are revised to be more interactive.
Tutorials. Tutorials in Introductory Physics (TIP) is a
supplementary curriculum developed at the University
of Washington (UW) [27]. It consists of a series of
pretests, worksheets, and homework assignments
intended to develop conceptual understanding and
qualitative reasoning skills. The tutorials are designed
for use in a small classroom environment in which
students work in groups of 3 or 4, but it can also be
adapted for larger class settings. The UW Tutorials
were developed from several interview-based studies
on student understanding of topics in velocity and
acceleration [269,270,286], kinematics graphs [287],
Newtons laws [288], energy and momentum
[289,290], geometric optics [31,291], and electricity
and magnetism [12,270]. More recent instructional
020119-16
strategies and materials have been developed

for topics in light and electromagnetic waves
[11,292294], gas laws [295,296], thermodynamics
[297,298], hydrostatic pressure [299], static equilibrium [300], and special cases in momentum [301].
The University of Maryland expanded upon the UW
tutorial framework to develop Activity-Based Tutorials
(ABT), which also relate concepts to mathematics and
include technological tools such as computers for data
acquisition and displaying videos or simulations
[225,226,302]. Another adaptation from University of
Maryland, called Open-Source Tutorials (OST), permits
instructors to modify worksheet materials to meet their
needs and places emphasis on developing both students
concepts and epistemologies or beliefs about learning
physics [224]. Studies of the effectiveness of TIP and
ABT tutorials have produced high gains on multiple-choice
questions and pre-post diagnostic measures such as the
Force and Motion Concept Evaluation, but sometimes with
mixed attitudinal reactions from students [303305].
Although students reported they viewed tutorials as useful
and they liked working in groups, very few (20%) said they
enjoy tutorials [303]. A comparison of the three types of
tutorials indicates that students using the OST version
perform better on course exams and diagnostic tests than
TIP or ABT [306].
Cooperative learning. Cooperative learning refers to a
strategy for collaborative group activities developed
by Johnson et al. [307,308]. The principles of cooperative learning have been applied to physics
classes in an approach called cooperative group
problem solving [110,119] and also integrated into
other approaches, such as Just-In-Time Teaching,
problem-based learning [236], and Workshop-StudioSCALE-UP physics. Cooperative group problem
solving necessitates the use of problems appropriate
for group work, such as context-rich problems [178]
or ill-structured problems [236] and has been shown to
improve problem-solving performance (for a comprehensive list of alternate problem types, see the Sec. III).
However, there are several structural aspects that
influence group functioning, such as the size of the
group (the optimal size is three), composition (heterogeneous by ability), seating arrangement, use of group
role assignments, and testing procedures that include
both group and individual components [110].
understanding during class (see the discussion of laboratory

reform in Ref. [205]). Also reviewed here are technological
tools used in labs.
Microcomputer-based labs. With the introduction of
microcomputer-based laboratory (MBL) tools such as
sonic rangers, it became possible for students to
observe a graph produced simultaneously with the
motion of an object. Research on these tools and
curricula developed to accompany them determined
that they improved students understanding of kinematics concepts and graphs (e.g., position, velocity,
and acceleration) in comparison to paper-and-pencil
activities [309,310]. Some researchers attribute the
success of MBLs to their real-time data collection, by
which students can observe both the motion of an
object and a graph of the motion generated simultaneously [309], whereas others do not see a difference
between real-time and delayed-time analysis of motion [311,312]. A curriculum that utilizes MBL
tools to address common student preconceptions is
RealTime Physics Active Learning Laboratories
[227230] and computer-based problem-solving labs
at the University of Minnesota (UMN), written in
LabVIEW [313]. Studies of the RealTime Physics
laboratories indicate improved student learning of
concepts in dynamics, as measured by performance
on the FMCE exam [314]. Studies of the early
implementation of UMN computer-based problemsolving labs indicated no significant differences
between groups using computers and groups using
traditional tools with respect to course grades or the
Test of Understanding Graphs in Kinematics (TUG-K)
inventory; however, the experimental (computer)
treatment group scored slightly better on the Force
Concept Inventory post-test [234].
Video analysis software. Another computer-based tool
that has been integrated into instruction is video
analysis software [231,315]. Students who had opportunities to collect and analyze data resulting from
the motion of objects in videos performed better on
interpreting kinematics graphs; a teacher demonstration alone was not sufficient for producing significant
results [231]. The software was also perceived as
useful by students and improved students comfort
with using computers [316].
Engaging students in the process of science. Some
instructional strategies for the laboratory seek to
engage students in the authentic practice of science
in open-ended experiments, such as observing new
phenomena, developing hypotheses, designing their
own experiments to test them, measurement and
uncertainty, and reporting results in written papers
[72,317,318]. One curriculum that emphasizes the
development of scientific abilities is the Investigative Science Learning Environment (ISLE) labs
3. Laboratory methods
These approaches seek to revise traditional labs, which
are often described as confirmatory or cookbook labs in
which students follow step-by-step instructions to verify a
result. General approaches to making labs more interactive
include engaging students in class discussions before and
at the end of class; revising materials such that students
must plan and make decisions; and probing for student
020119-17
[233,317,319,320]. ISLE has been shown to develop

students scientific abilities, such as multiple representations, designing an experiment, collecting and
analyzing data, evaluating an experimental outcome,
and scientistlike communication [233,272,321].
Students who participated in physics laboratory
design activities were also able to transfer skills
such as evaluating assumptions and experimental
uncertainties to a novel task in biology [233]. In
ISLE students go through a learning cycle of
observation and exploration, developing explanations, designing experiments to test out and refine
their hypotheses, and applying knowledge to solve
problems [233,266]. Although labs are a key component of the ISLE curriculum, it also includes
worksheets and activities appropriate for recitations
and large class meetings in the Physics Active
Learning Guide [319].
Another laboratory curriculum is the Problem Solving
Labs developed at the University of Minnesota [322]. Each
laboratory experiment is motivated by a context-rich
problem, and students are required to complete guiding
warm-up questions to reach a prediction before attending
the lab session. During lab they discuss predictions in their
cooperative groups, and are prompted to make decisions
such as planning the data collection and analysis. Although
there are research results with respect to UMNs contextrich problems and cooperative group problem solving
[110,119], in addition to a comparison of computer and
noncomputer labs [234], research on the effectiveness of
the problem-solving laboratory curricula itself is not
currently available and warrants future study.
A laboratory curriculum called Scientific Community
Labs was developed at the University of Maryland [323].
The labs take a conceptual approach to learning measurement techniques, including the concept of uncertainty
[235]. During lab, students are given a question and must
devise an experiment, collect and analyze data, and present
their results to the class. Confidence in the results (or error
analysis) is an explicit part of the discussion. Students
using the reformed laboratories exhibited higher gains on a
physics measurement questionnaire and more sophisticated
reasoning when comparing data sets, such as using range
overlap rather than percent difference [235].
4. Structural changes to the classroom environment
This section reviews instructional interventions that alter
the standard classroom structure. In some instances this
includes changing the lecture seating and incorporating
technology to accommodate group work on meaningful
problems, and in other situations the entire course is
restructured to combine lecture, lab, and recitation into a
single interactive environment (e.g., workshop and studio
class settings).
Problem-based learning (PBL). Developed at the

University of Delaware in the early 1990s, problembased learning [324] has been implemented in a
variety of undergraduate science courses, including
some in physics [236]. Students work in small
cooperative learning teams to solve a complex, realworld problem that may take several weeks to investigate. Lectures are kept to a minimum, and the
instructor and/or peer tutors help facilitate the functioning of groups and ask probing questions during
class. Research on PBL in physics classes has determined that it helps students develop critical thinking
and communication skills [325,326].
Physics by Inquiry (PbI). Developed at the University
of Washington, the Physics by Inquiry curriculum
includes a workbook of narratives, laboratory-based
experiments and exercises, and supplementary problems [238,239,327]. Students keep a notebook of their
observations and exercises, to help them reflect on
how their understanding of science concepts changes
throughout a course. The targeted audiences for PbI
include preservice and in-service K-12 science teachers, underprepared students, and nonscience students.
Many of the research studies to develop and evaluate
the PbI materials were conducted concurrently with
the studies of Tutorials in Introductory Physics
described above. Examples include curricula on light
and shadows [31] or on density (e.g., sinking and
floating) [328]. A study comparing Physics by Inquiry
to traditional instruction found that elementary education students in a PbI course performed significantly
better on a quantitative and a qualitative exam problem
than physics and engineering students in a traditional
introductory physics course [329].
Workshop Physics (WP) and Explorations in
Physics (EiP). Workshop Physics was developed at
Dickinson College, where the calculus-based physics
course was restructured to eliminate formal lectures
[240,241,330]. Instead, students meet for three 2-hour
sessions per week to engage in experiments and
analyze data using computer tools, such as curve
fitting with spreadsheet software. Research on WP
indicates that students prefer the workshop method,
have improved performance on conceptual questions
(at some but not all institutions), and equal problemsolving performance in comparison to traditional
sections of the same course [240]. The WP activities
are published in multiple workbook volumes, called
the Workshop Physics Activity Guide [241]. EiP is the
equivalent of Workshop Physics, but designed for
nonscience students [331].
Studio Physics. The term studio comes from
Wilsons description of a comprehensive unified
physics learning environment or CUPLE physics
studio implemented at Rensselaer [246]. In Studio
020119-18
Physics, the lecture and laboratory sessions are combined into a single course session of 3045 students
meeting for two hours, 24 times per week [242] with
computer-based activities and collaborative group
work. Early implementations of Studio Physics
showed low gains on the Force Concept Inventory
exam (similar to scores from traditional courses), but
the introduction of research-based methods such as
interactive lecture demonstrations and cooperative
group problem solving significantly improved learning gains [242]. Hoellwarth, Moelter, and Knight
[243] found that students in a studio section had
significantly higher normalized gain on the FCI and
the FMCE than students in a traditional classroom;
however, students scores on quantitative final exam
problems were the same or slightly worse in the studio
sections.
In another instantiation of studio, called New Studio
Physics, the lecture remains separate but the recitation and
laboratory sessions are combined into a 2-hour session, two
times per week [245]. Implementation of New Studio
Physics found gains on the Force Concept Inventory exam
similar to those obtained by other interactive engagement
instructional methods.
SCALE-UP. The acronym SCALE-UP stands for
student-centered active learning environment for
undergraduate programs, which was first developed
and studied at North Carolina State University in
physics classes and has been adopted by several
other programs and institutions [247,332]. SCALEUP uses a very carefully designed classroom in which
students (typically 50100) work at round tables in
teams (typically 3 teams of 3 students at each table)
with networked laptops and access to whiteboards.
There are multiple display screens around the room.
Students engage in hands-on activities including viewing interactive computer simulations, responding to
questions or problems, and conducting hypothesisdriven laboratory experiments. Research on SCALEUP cites improved problem-solving scores on course
exams, conceptual understanding gains as measured by
pre-post concept inventories, slightly more positive
attitudes, and reduced attrition (see How do you know
it works? in Ref. [333], and see also [247,334]).
TEAL. One extension of Studio-SCALE-UP is the
Technology-Enabled Active Learning (TEAL)
project at the Massachusetts Institute of Technology
[248,250]. TEAL uses the classroom design of
SCALE-UP, which combines lecture, recitation, and
laboratories into a single media-enhanced classroom.
Like SCALE-UP, it incorporates Web-based homework assignments, conceptual questions with clickers,
and two- and three-dimensional visualization tools for
learning electromagnetism concepts [250]. Studies
of TEAL indicate higher learning gains on pre- and
post-tests of multiple-choice conceptual questions

(which were adapted from a variety of research-based
sources) and a lower failure rate than the traditional
course structure [249,335]. Open-ended surveys and
focus groups indicated that students attitudes toward
the TEAL classrooms varied, with several students
expressing positive comments regarding the usefulness of the collaborative activities and others expressing they preferred a lecture format, or thought that
group work was too unguided [249].
5. General instructional strategies and materials
Textbooks. Understanding Physics is a text written by
Cummings et al. [251] based on results from physics
education research. The text includes touchstone
example problems and alternate types of problems,
microcomputer-based laboratory tools, and topic
organization to facilitate student understanding. The
text Physics for Scientists and Engineers: A Strategic
Approach by Knight [336] also incorporates PERbased ideas such as multiple representations, alternate
problem types, and a four-step problem-solving strategy for all worked examples: model, visualize, solve,
and assess. The curriculum Six Ideas that Shaped
Physics was developed and tested over a period of
more than eight years as part of the Introductory
University Physics Project (IUPP) at Pomona College
[337]. The volumes are structured to emphasize the
hierarchical structure of physics (e.g., volume C on
conservation laws) and provide support for active
learning. Courses using the materials found very
high gains on conceptual inventories including the
FCI and Brief Electricity and Magnetism Assessment
(BEMA), with normalized gains in the range of
0.500.70 (see Evidence for Success, in Ref. [338]).
Another research-based curricula is Matter &
Interactions (M&I) developed at North Carolina State
University [252]. M&I approaches calculus-based introductory physics from a modern perspective, highlighting
the atomic nature of matter, relativity, and computational
modeling of complex systems [339341]. Matter &
Interactions has resulted in higher gains on the BEMA
pre-post diagnostic test [342], but lower gains on the FCI
compared to traditional instructiona finding that perhaps
could be explained by differentially more coverage of FCI
topics in the traditional curriculum [343].
Problem solving. For a review of instructional strategies and materials for teaching problem solving, see
Sec. III. The methods reviewed there include alternate
types of problems, qualitative approaches, teaching an
explicit framework, and Web-based homework and
computer tutors for problem solving.
Animations, simulations, and computer programming.
A recent trend in instructional strategies and materials
is the use of technological tools in the classroom, such
020119-19
as animations, simulations, visualization tools, and

computer programming experiences. With respect to
research on animations, Dancy and Beichner [344]
compared students responses to Force Concept
Inventory questions with static pictures and computer
animations for the same questions. They found that the
students were better able to interpret the questions
intention when viewed in animated form, and they
gave an answer more diagnostic of their understanding. Another study by Yeo et al. [345] found that
students using interactive media often retained their
intuitive (incorrect) conceptions about motion unless
they had explicit guidance from researcher interventions. Steinberg [346] did not find performance
differences between students who interacted with a
simulation for learning air resistance compared to
those who completed paper-and-pencil versions; however, students who interacted with computer simulations were more reliant upon the computer to provide
them with the correct answer. Computer visualization
tools have been shown to be particularly effective
for abstract concepts in electricity and magnetism,
such as the concept of a field [217,218,249,335,347].
Although textbooks often include access to online
visualization or animation tools, research support for
their development and effectiveness is not always
available and/or cited [348].
One of the most widely cited projects to develop and test
interactive science simulations is the University of
Colorados PhET project [349]. The Web site includes a
range of free simulations of phenomena for physics,
biology, chemistry, Earth science, and math; it also includes
a bank of activities created by teachers to accompany the
simulations. The physics section includes (to date) more
than 70 simulations. There have been several research
studies published regarding the effectiveness of the PhET
simulations [253,254,350]. One study in particular [255]
found that students who used a computer simulation to
explore direct-current circuits rather than real light bulbs,
meters, and wires had higher scores on their lab report,
scored higher on final exam questions pertaining to circuits,
and were more successful at performing tasks including
assembling a real circuit. They cite that computer simulations are particularly useful for instruction in that they
allow students to view hidden phenomena, such as
current flow, and focus their attention on relevant details
of the representation.
Finally, a few studies have explored the use of computer
programming in physics courses. An early study with
programming in a course for physics majors was the
project M.U.P.P.E.T.: Maryland University Project in
Physics and Educational Technology [257]. The
M.U.P.P.E.T. project determined that programming in
introductory physics facilitated the use of more realistic
problems and modern topics, and the use of carefully
designed projects could encourage students use of qualitative and analytic reasoning skills. Another instantiation of
computer programming in introductory physics is the use
of VPython in the Matter & Interactions curriculum [256].
In addition to teaching students some basic programming
and debugging skills, the computational tools also allow
students to visualize abstract phenomena and tackle
problems that cannot be solved analytically. For a more
comprehensive review of research on integrating computational physics into the undergraduate curriculum, see
Ref. [351].
Homework. There are few studies related to effective
procedures for the assignment, collection, and grading of
homework in large enrollment courses. One study found
that allowing students some choice in which problems to
solve and making a fraction of problem solutions available
before homework is due provided more timely feedback to
students, and resulted in an increase in self-direction [352].
Preservice and in-service teachers. As mentioned
previously, the Physics by Inquiry laboratory materials
are appropriate for educating preservice and in-service
K-12 teachers about physical science [238,239]. A
different curriculum targeted toward elementary
teachers is Physics and Everyday Thinking (PET)
[353]. PET has shown a shift toward expertlike
attitudes as measured by the Colorado Learning
Attitudes about Science Survey [354].
High school curricula. Although not reviewed here, it
is worth mentioning that there have been several texts
and workbooks developed for use in middle school
physical science and high school physics, which are
sometimes adapted for undergraduate courses with
nonscience majors. Examples include Modeling
Instruction [258,260], Minds on Physics [259], Its
About Time: Active Physics [355], Tools for Scientific
Thinking [239], and a curriculum for 7th-8th graders
called InterActions in Physical Science [356].
Upper-division curricula. One curriculum that reforms the junior-level and senior-level courses for
physics majors is the Paradigms in Physics project at
Oregon State University [268,357]. It includes revisions to the structure or order of topics, content of
courses, and instructional methods, in an effort to
better reflect physicists views and experiences in the
field. Examples include adding laboratory experiments as a component in several courses, making a
stronger link between mathematics and physics, and
introducing computational examples and problems
[357]. Evaluation of the program indicates improved
retention of physics majors and improved physics
problem-solving abilities [268]. The development
and evaluation of UW Tutorials curricula for upperdivision mechanics courses also shows promising
effects on learning advanced topics [358]. The
University of Colorado-Boulder has transformed their
020119-20
upper-division electricity and magnetism course

[359362] and quantum mechanics course [363] by
integrating active engagement methods such as using
conceptual questions during lecture, tutorials during
recitations, and homework that utilizes multiple representations. They have identified several areas in
which upper-division students have difficulty, such as
communicating understanding, visualizing fields,
problem-solving techniques [359], and the use of
conceptual clicker questions can give faculty insight
into students understanding [362].
Reform to graduate education is less common. Carr and
McKagan [364] researched the implementation of PER
methods into a graduate quantum mechanics course,
including changes to the content, textbook, instructional
methods, and assessment. Materials were adapted during
the course in response to student surveys and interviews.
They found that students became more interested in course
material when the instructor made explicit connections to
modern research, and students were responsive and positive
toward the lecture modifications, such as the increase in
peer discussions during class.
The learning assistant (LA) model. A curricular
reform developed at the University of Colorado [365]
uses undergraduate students who performed well in an
introductory course to serve as learning assistants in
following years. LAs facilitate small group interactions in lectures and recitation sections. Use of
LAs in teaching has resulted in higher learning gains
for students in the BEMA and FMCE than achieved
with traditional instruction; serving as a learning
assistant also increases dramatically the likelihood
that they will enter the teaching profession. Another
study indicates that LAs help improve student performance when used in lab settings [366].
Although many of the instructional materials described
here are stand-alone curricula, they are sometimes implemented in combination. Redishs [205] text outlines how
the Physics Suite integrates several curricula into an
activity-based physics program, including Understanding
Physics [251], alternate types of problems, interactive
lecture demonstrations [214], Workshop Physics [241],
and computer tools in workshops and labs with
RealTime Physics [227230].
of new faculty workshops sponsored by the American

Association of Physics Teachers, American Physical
Society, and American Astronomical Society [367].
E. Strengths and Limitations

1. Strengths
A strength of research on curriculum and instruction in
physics is the sheer abundance of methods that have been
developed and evaluated with research. In particular, many
early approaches have been tested and refined over several
decades to the point of becoming mainstream in introductory undergraduate courses (such as the use of conceptual
questions and polling technologies, like clickers). Adoption
of research-based instruction is also spreading as a result
2. Limitations
This review identified some instructional strategies and
curricula that have not been extensively tested for their
effectiveness in physics, such as the Just-in-Time teaching
method and several laboratory-based curricula. In addition,
although some textbooks are PER based, many of the most
commonly used texts for undergraduate courses are not,
and this can create a mismatch between the goals of
reformed instruction and the materials actually used in
the course; to date, a large percentage of college physics
courses are inconsistent with evidence-based reforms
[368,369].
Upper-division undergraduate courses and graduate
education. An area of future study includes expanding
evidence-based instructional strategies and materials
that have been developed for introductory undergraduate physics courses to other populations, such
as upper-division (junior- and senior-level) courses
for physics majors and courses for graduate students.
Although there have been some studies of curriculum and instruction in upper-division courses
[268,362,364,370], this area would benefit from additional research.
High schooluniversity cross-curricular studies.
Although several of the physics curricula are identified as appropriate for a particular student population,
it is possible that some curricula for high school might
be appropriate for university nonscience majors (see
Ref. [260]) and that university curricula might be
appropriate for advanced high school courses. The
application of instructional strategies and materials to
alternate groups of students is a possible area for
future research.
Courses for biology and premedicine students. There
is some current interest in reforming the introductory
course for biology and premedicine students [371].
Additional discussion is needed to determine how
current courses are (or are not) meeting the needs of
this student population, and research is needed to
develop and test reformed instruction and curricula. At
least one research group in Germany has developed
and tested the effectiveness of a reformed physics
laboratory curriculum for medical students [372].
They found that including medical applications
(e.g., connecting electricity and neural functions)
improved students attitudes towards physics and
improved their ability to relate concepts of physics
and medicine as measured by the construction of
concept maps.
020119-21
The influence of instructor modifications on the

success of reformed curricula. Instructors do not
always implement PER-based instructional methods
in the way that the developers intended them, and it is
unclear what effect this has on the effectiveness of the
method. An area of current research is to explore ways
in which instructors modify methods when implementing them in their classes, their reasoning for
doing so, and how this influences student learning. For
examples of some studies, see Sec. VII.
Lab instruction. Much of the early research on labs
focused on the integration of computer-based sensors
and technological tools into the laboratory, but it is
possible that many institutions still use confirmatorystyle, cookbook labs, despite the use of reformed
instruction in other aspects of the course (such as
lecture and recitation). With the exceptions of ISLE,
RealTime Physics, and integrated course structures
(workshop, studio, SCALE-UP, and TEAL), there has
not been much published research on the effectiveness
of laboratory curricula.
Textbooks. Although there is unpublished research
about the readability of introductory physics textbooks [373], there is a need for more research on how
students read and interact with physics textbooks and
components in differing formats, such as equations,
diagrams, text, colored photos, etc. Cognitive load
theory suggests that many enhancements to textbooks may create extraneous memory load that detract
from, rather than add to, learning, although little
research exists on the design of effective textbooks
(see Refs. [374,375] for reviews).
Instructional technology. Although there were several
aspects of technology identified here (e.g., classroom
polling technologies, multimedia presentations, computer data collection and analysis tools, and computer
programming), it is expected that technology and
computer-based instruction will continue to evolve,
and this will require future research on the effective
use of these technologies in classrooms.
V. ASSESSMENT
The development and validation of multiple-choice
concept inventories was (and continues to be) an influential
area of research in physics education, particularly the
development of mechanics and force concept tests, such
as the Mechanics Diagnostic Test [376], Mechanics
Baseline Test [14], the FCI [15], and the FMCE [17]. In
particular, the FCI has experienced widespread use to
evaluate the effectiveness of instruction within and across
several institutions [209]. The bank of available concept
inventories has expanded to more than 30 physics inventories, including surveys of attitudes and beliefs [377].
The process for developing and validating concept tests
typically begins with qualitative studies of student
difficulties and incorrect responses (or a review of existing

research on common misconceptions), drafting and piloting
questions, and a series of statistical procedures to evaluate
items and scores (see the handbook [378]; see also [379]).
For a comparison across several different inventory development methodologies, see Lindell, Peak, and Foster [380].
Research on assessment has broadened to include exploring
correlations between inventory scores and other measures
of performance, comparing scores across multiple populations (culture and gender), and exploring the value of
complex models of student learning beyond pre-post
scores. Assessment also refers to alternate formats for
measuring student understanding such as rubrics and
exams, which will also be discussed in this section.
1. Development and validation of concept inventories
To what extent do scores on a multiple-choice concept
test agree with students written responses to open-ended
questions (a measure of test score validity)? How does
the context and format of concept test questions impact
student responses? How do the conditions under which a
concept test is administered affect students performance
on the test? This area of research includes developing
and validating concept inventories, methods for analyzing
results from concept inventories [381], and issues related to
context sensitivity and studies of testing conditions [382].
2. Comparing scores across multiple measures
What is the relationship between students scores on the
Force Concept Inventory and performance in a physics
course as measured by grades? What is the relationship
between students performance on concept tests and
measures of their mathematical skills? Several studies
have compared scores on concept inventories to other
measures such as SAT scores [383], scientific reasoning
ability [384], scores on a math diagnostic test [385], as
well as how students blend mathematical and conceptual
knowledge [386]. Other possible measures of comparison
include grade point average, course grades, or final exam
scores. Some studies also compare tests on one inventory
to those on another, such as comparing the FCI to the
FMCE [387].
3. Comparing scores across multiple populations
(culture and gender)
How do females average scores on the Force Concept
Inventory test compare to males average scores? How do
FCI scores compare across cultures (e.g., American
and Chinese students)? Are there different gender
patterns in clicker responses? There have also been some
recent studies comparing inventory scores across multiple
social or cultural groups, such as comparing American
020119-22
and Chinese students [388,389] and comparing males

and females [390392]. There is also some research on
gender differences in the use of online homework [393]
and in the response patterns of males and females using
clickers [394].
The most recent edition of the Standards for Educational

and Psychological Testing describes validity in the following way:
4. Course exams and homework

To what extent does performance on multiple-choice
questions agree with students performance on equivalent
open-ended problems? How frequently should homework
be assigned and graded, if at all? What format of grader
feedback on homework is most effective for promoting
student learning? This area includes research on multiplechoice course exams, particularly their equivalence to
performance on open-ended problems [395,396] and homework grading [352].
5. Rubrics for process assessment
To what extent do scores on rubric-based grading agree
with other measures of student performance (e.g., traditional grading of a problem solution)? What is the level of
agreement (consistency) among multiple people using a
rubric to score the same student papers? Research questions that pertain to the development of rubrics focus on
the validity and reliability of scores on the assessment.
Examples of physics rubrics include the scientific abilities
rubrics developed for the ISLE curricula [317] and problem-solving rubrics [205,397]. There are also assessments
under development for problem solving [398].
6. Complex models of student learning
To what extent do predictions from mathematical models
agree with measures of student learning? How are student
responses to conceptual questions affected by instruction?
What is the pattern of student learning throughout a
course? Some researchers are attempting to create mathematical models to describe student learning [399,400] and
others are investigating the peaks and decays of learning
that occur between pre- and post-testing [401,402].
Research on the development and validation of assessments is typically guided by theories of test development
from fields such as quantitative measurement in education
(see Ref. [378] for an introduction to classical test theory;
see also Ref. [403]. It is also guided by evolving definitions
of reliability and validity (see Chap. 3 of Ref. [205] for a
review). Historically, validity was defined as determining
whether an assessment measures what it claims to measure
and relied on statistical measures for each type of
validity, such as content, construct, predictive, and concurrent validity. The current view of validity has shifted
away from the types paradigm to a unified concept of
validity that includes multiple sources of evidence [404].
Validity refers to the degree to which evidence and

theory support the interpretations of test scores entailed
by proposed uses of tests The process of validation
involves accumulating evidence to provide a sound
scientific basis for the proposed score interpretations.
It is the interpretations of test scores required by
proposed uses that are evaluated, not the test itself.
(p. 9).
The 1999 edition of the Standards for Educational and
Psychological Testing cites five primary sources of evidence for validity: evidence based on test content, response
processes, internal structure, relations to other variables
(external structure), and consequences of testing. For
example, the aspect of content considers whether the tasks
or items on an assessment are representative and relevant to
the domain or topic being tested. The aspect of response
processes considers whether the processes engaged in by a
respondent while taking a test are consistent with the
assessment developers intentions. Internal and external
structure aspects consider interrelationships among parts of
a test and comparisons with other measures of the same
construct.
In addition to issues of validity, the Standards for
Educational and Psychological Testing also address the
importance of reliability and utility for designing tests.
Reliability refers to the consistency or stability of scores
and score interpretations on an assessment, such as agreement among multiple graders for a free-response question.
Utility refers to the potential usefulness of the assessment as perceived by its users, such as researchers and
instructors.
The development of assessments is also guided by
research literature on test design, such as the procedures
outlined by Engelhardt [378] and Treagust [379].
According to these sources, a general procedure for
developing concept tests in science includes the following
steps: (1) identify knowledge or define the content area in
science you want to test; (2) identify student misconceptions associated with that topic by examining the existing
literature, or by student interviews and responses to openended questions; (3) develop items and pilot test them in
two-tier formasking for response and written reasoning
for the response.
and data analysis)
1. Contexts
Basic research to identify common student difficulties is
typically done outside the classroom, in an experimental or
clinical setting with paid volunteers (refer Sec. II). During
020119-23
the development of concept assessments, sample items may

also be pilot tested with student volunteers, or could be
tested in the context of a course. Statistical analyses of
quantitative assessment scores are typically conducted in
conjunction with a course where the data collected includes
student responses to concept inventory questions, homework, or exams.
validity of those categories, prompting a discussion

regarding what the FCI actually measures [405407].
FMCE: Force and Motion Conceptual Evaluation
[17]. An analysis by Ramlo [408] established validity
and reliability estimates for the FMCE.
TUG-K: Test of Understanding Graphs in
Kinematics [13]
CSEM: Conceptual Survey in Electricity and
Magnetism [16]
ECS: Energy Concepts Survey [409]
DIRECT: Determining and Interpreting Resistive
Electric Circuits concept Test [410]
BEMA: Brief Electricity and Magnetism Assessment
[411]
R-FCI: Representational variant of the Force Concept
Inventory [412]
QMCS: Quantum Mechanics Conceptual Survey
[413]
CUE: Colorado Upper-Division Electrostatic
assessment [414]
FVA: Force, Velocity and Acceleration assessment
[415]
Other assessments:
LCTSR: Lawson classroom test of scientific reasoning [416] (reproduced in Coletta and Phillips [384]).
This assessment includes questions on conservation, proportional thinking, identification of variables,
probabilistic thinking, and hypothetico-deductive
reasoning.
Statistical analysis of scores. In a meta-analysis of
FCI scores, Hake [209] uses a measure he refers to as
normalized gain to compare pretest and post-test
scores, which is examined in more detail in Ref. [417].
The average normalized gain for a group of students is
computed by taking the average percent gain (average
post-test score %average pretest score %) divided
by the possible gain (100%average pretest score %).
Hakes analysis concluded that courses utilizing interactive engagement methods produced an average
normalized gain of 0.48 0.14 standard deviations on
the FCI, whereas courses taught using traditional
methods produced an average normalized gain of
0.23 0.04 standard deviations. It has become a
common practice in PER to compare the normalized
gain values obtained in a study to these values reported
by Hake [209].
Although the normalized gain measure is popular, it has
received some criticism. Marx and Cummings [418]
suggest revising the gain measure to a proposed measure
of normalized change. The main distinction between the
two measures is that normalized change revises the formula
for situations in which the post-test score is lower than the
pretest score, and drops students who score 100% on the
pretest (a normalized gain of infinity).
Ding and Beichner [381] identify five statistical means
for analyzing test scores to assess the validity and reliability
2. Participants
Studies of physics assessments frequently involve
(algebra based or calculus based) at a single institution over
one or two semesters, sometimes longer. Studies conducted
with high school students or disciplinary majors are less
common. Very few meta-analyses have been conducted to
compare concept inventory scores across multiple institutions. One such analysis is Hake [209], who examined
Force Concept Inventory scores for 62 mechanics courses,
comparing interactive-engagement instructional strategies
to traditional instructional methods.
The early development stages of a concept inventory
typically include interviewing students or reviewing
existing research literature to identify misconceptions
and/or propose assessment items, and pilot testing these
items in both a free-response format and a multiple-choice
format. After an assessment has been drafted, the statistical
analysis of scores (often to measure the validity and
reliability of scores on items) typically involves numerical
data and quantitative data analysis. Studies comparing
performance on multiple measures or across student populations typically utilize statistical measures, such as
correlation coefficients.
D. Findings
1. Development and validation of concept inventories
Concept inventories used in physics. As stated previously, more than 30 physics concept inventories or attitude
surveys have been written [377]. Many are still under
development and undergoing validation procedures, and
about half of them have been published. The inventories
with accessible references are listed below, in chronological
order of their publication date.
Published concept inventories and validation studies:
MDT: Mechanics Diagnostic Test [376]
MBT: Mechanics Baseline Test [14]
FCI: Force Concept Inventory [15]. In the initial
article published about the FCI, they identified six
conceptual dimensions included on the test (kinematics, Newtons three laws, the superposition principle,
and kinds of force). These came into question when
factor analysis could not statistically confirm the
020119-24
of those scores: classical test theory, factor analysis, cluster

analysis, item response theory, and model analysis.
Other publications explore these and other analysis methods in greater detail, including item response theory
[121,419,420], cluster analysis [421,422], Rasch model
based analysis [423], concentration analysis [424], and
model analysis [399].
Influence of testing conditions. A study by Ding et al.
[382] outlined several testing conditions that affected
pre- and postinstruction scores on the Conceptual
Survey in Electricity & Magnetism (CSEM) exam.
For example, they found that moving the test administration time by a few days or lectures could have a
substantial effect on the CSEM pretest scores, since
topics of charge and Coulomb force are often discussed in the first week of the course. This is an
important consideration for situations in which concept tests are administered outside of a lecture class,
such as during a laboratory or recitation section. They
also investigated the effect of several different incentives on post-test scores: giving a few points to take
the post-test regardless of performance did not seem to
influence overall class average; giving students the
option to replace an exam score with their post-test
CSEM score resulted in a lower response rate but
increased average post-test score (perhaps due to
sampling from a higher-achieving group of students),
and incorporating CSEM questions into the final exam
produced a significant increase in post-test scores
[383]. Henderson [425] observed only a small increase in scores when the FCI post-test was graded
versus when it was ungraded, a difference of about
half an item or 2%.
Web-based assessments. Dancy and Beichner [344]
compared performance on a paper-and-pencil version
of the FCI to an animated version on the computer.
They found that students were better able to understand the intent of the question when viewing it in
animated format and this version was more indicative
of their conceptual understanding.
Bonham [426] studied the reliability, compliance,
and security of using Web-based assessments in an introductory astronomy course. The study found no statistical
differences between administering an astronomy concept
inventory and an attitude survey on paper or on the Web.
The compliance rate increased with the amount of course
credit given for completing Web-based assessments and
the frequency of Email reminders. Most students abided
by the request not to print, save, or copy portions of the
assessments; a small number of students navigated away
from the assessment to use other Web-based applications,
suggesting they might have used Web-based resources. The
study was unable to detect the use of non-Web resources
for completing assessments (such as textbook, notes, or
classmate).
Sensitivity to question context. Stewart, Griffin, and

Stewart [427] studied the effect of changing question
context on the FCI on performance. They investigated
eight types of question alterations: changing a concrete system to abstract (like changing a person
pushing on a box to a force exerted on a block),
changing the objects in the question to other realistic
objects, removing redundant wrong answers, adding a
picture, removing a picture, reordering multiplechoice answers, restructuring the order of a group
of questions, and making responses symbolic instead
of textual. The strongest positive effects were seen for
removing a figure, changing the objects in the question, and removing questions from a group. Although
altering a question sometimes vastly affected responses on that particular question, these effects
tended to balance each other out and resulted in no
overall significant change to the total score.
Steinberg and Sabella [50] report differences in students
answers when a FCI question is administered in multiplechoice format to when it is an open-ended question on an
exam. Students typically scored higher on the exam
questions, which were less colloquial than the FCI and
often prompted them to draw the forces acting on objects.
Although they observed a correlation between student
performance on the two formats, there were several
instances in which students gave inconsistent answers.
The researchers caution educators to be wary that student
understanding is more complex than simple correct or
incorrect answers to multiple-choice questions, since
changing the context of the test can alter responses.
Dufresne, Leonard, and Gerace [49] also reported specific
examples of how changing FCI questions alters students
answers. They describe one example in detail: the situation
of two balls of unequal weight dropped from the same
height. Although the majority of students surveyed
responded correctly to the FCI question that the balls
would reach the ground at about the same time, several of
these students incorrectly believed that the forces acting on
the balls were also equal. The possible influence of context
familiarity was briefly addressed by Huffman and Heller
[407], who suggested that the degree of familiarity with
particular objects, such as hockey pucks or rockets, might
affect a students scores on those items. Context sensitivity
has been found in other studies of physics concepts as
well [428].
2. Comparing scores across multiple measures
Comparing inventory scores to other measures. The
widespread availability of physics concept inventories
has prompted some researchers to question whether
these tests can be used as placement exams. In other
words, is a students performance on a test like the FCI
a predictor of their performance in an introductory
physics course? Are strong math skills essential for
020119-25
learning physics? Several studies have compared

scores on physics concept inventories to other performance measures such as SAT scores, grade point
average, course grades, exam scores, math tests,
problem solving, and Lawsons classroom test of
scientific reasoning. Meltzer [385] reported significant
correlations between normalized gain on a Conceptual
Survey of Electricity (CSE) with pretest scores on an
algebra or trigonometry math skills test for three out of
four groups (r 0.38, 0.46, and 0.30). Coletta,
Phillips, and Steinert [383] reported strong correlations between normalized gain on the FCI and SAT
scores (r 0.57 at a high school and r 0.46 at a
university), and correlations between normalized gain
on the FCI and scores on Lawsons classroom test of
scientific reasoning (r 0.51). Another study claims
that math ability is a significant predictor of performance in a physics course [429,430]. Henderson [425]
emphasizes that the FCI cannot be used as a placement
test, because although high-achieving students
(scoring higher than 60% on the pretest) usually get
an A or B in the course, the test does not do a good job
of predicting failure.
Comparing scores on multiple physics inventories.
One study has compared scores on the FCI to scores on
the FMCE, which both cover one-dimensional kinematics and Newtons laws, but measure slightly
different content and differ in format [387]. For
example, the FMCE includes questions in pictorial
and graphical format whereas the FCI is primarily in
verbal and pictorial format, and the FCI is broader in
that it includes some questions on two-dimensional
motion with constant acceleration. The researchers
reported that scores on the FCI were significantly
higher than on the FMCE, and there was a strong
correlation between scores on the two exams, even
though some students showed inconsistent performance on the two tests. They also observed differences
in gains on the FCI and FCME for different instructional strategies. For example, cooperative group problem solving showed equal gains on both tests, whereas
using interactive lecture demonstrations showed higher
gains on the FMCE than on the FCI.
There is also some inconsistency in observing correlations between diagnostic pretest scores and normalized gain.
Hake [209] reported no correlation between FCI pretest
scores and normalized gain, and Meltzer [383] reported no
correlation between CSE pretest scores and normalized
gain, whereas Coletta et al. [383] did observe a correlation
between FCI pretest scores and normalized gain.
3. Comparing scores across multiple populations
Cross-cultural comparisons of concept inventory
scores. A comparison of concept inventory scores
for Chinese and American students found that Chinese
students scored higher on content knowledge tests like

the FCI and BEMA, but scored at the same level
as Americans on the Lawson Classroom Test of
Scientific Reasoning [388,389].
Gender comparisons of concept inventory scores. In
introductory calculus-based physics courses for scientists and engineers at large-enrollment institutions,
females are underrepresented and typically make up
20%25% of the class [390,431]. Recent analyses of
test scores in these courses indicate men significantly
outperform women on some physics concept inventories, such as the FCI and FMCE, by an average of
10%15%. There are conflicting views from different
institutions regarding the extent to which interactive
engagement instructional strategies can reduce and/or
eliminate the gender gap in scores. Lorenzo, Crouch,
and Mazur [391] claim that interactive engagement
instructional strategies such as Peer Instruction at
Harvard reduce or eliminate the preinstruction gender
gap in FCI scores. In partially interactive classes the
gender gap decreased from 13% before instruction to
7.8% after instruction, and in fully interactive teaching approaches it was essentially eliminated, decreasing from 9.2% to 2.4% [391]. In contrast, the
University of Minnesota observed a persistence of
the FCI gender gap on both pretest and post-test for
cooperative group problem solving instruction [431].
The data examined included 40 fall term classes
with more than 5500 students over ten years,
showing a gap of 15.3% 0.5% that decreased
slightly to 13.4% 0.6% although there was essentially no gender difference in course performance as
determined by course grade (1.5% 0.2%). The
University of Colorado at Boulder observed persistence in the gender gap on FMCE scores from pretest
to post-test but there were variations by instructor to
the extent curricula were implemented, such as
ConcepTests, online homework, and Tutorials in
Introductory Physics [359]. In partially interactive
classes, the gender gap averaged 10.9% preinstruction and 13.6% after instruction; in fully interactive,
it decreased slightly from an estimated 11.2% to
10.0%. In a second-semester course on electricity
and magnetism, there was a small gender gap on the
BEMA pretest (1.8% 0.5%), which increased to
6.1% 1.0% on the post-test. There were some
differences in course performance, where females
outscored males on homework and participation,
whereas males scored higher on exams, resulting
in overall similar course grades [390,392]. Kost et al.
[390] used logistic regression analysis to determine
that the odds of scoring above 60% on the post-test
were not different for males and females when
controlling for pretest scores (FMCE, math, and
attitude), suggesting that preparation and background
are responsible for the persistence of the gap, not
gender explicitly.
020119-26
Some researchers have attempted to account for pretest

gender differences by examining the preparation of males
and females entering introductory college physics courses
after high school. Kost, Pollock, and Finkelstein [390]
found that although there is a correlation between years of
high school physics and concept inventory post-test scores,
gender differences could not be explained by whether
students took high school physics because the gender
gap was present for matched groups. Hazari, Tai, and
Sadler [432] also found that background factors such as
math SAT scores, calculus, and high school grades were
equally predictive of success in university physics for both
males and females. Factors that differentially affected male
and female performance included high school pedagogy,
assessment characteristics, and affective factors such as
family attitude toward science and the fathers encouragement [432434]. Kost-Smith, Pollock, and Finkelstein
[435] suggest that the underrepresentation of women in
physics careers may be due to an accumulation of small
gender differences over time in aspects such as retention,
performance, and attitudes and beliefs such as personal
interest (p. 1). A recent intervention involving a brief
written exercise on the topic of values affirmation
successfully reduced the gender gap on FMCE scores at
the University of Colorado, suggesting that psychological
interventions hold promise for addressing performance
differences in science [436].
Gender comparisons on homework. Kortemeyer et al.
[120] determined that female students benefited more
from the online homework system CAPA than male
students as measured by final course grade. Further
study of gender differences in use of the CAPA
homework system found that more females than males
reported that they consulted peers online or discussed
problems with teaching assistants prior to attempting a
problem (21% females compared to 9% males),
whereas a higher fraction of men reported immediately attempting a problem after reading it (58% males
compared to 39% females) [393]. There were no
significant differences with respect to the time students reported spending on homework (females averaged 5.6 hours per week, males 5.1 hours per week).
Males and females differed in the nature of their
comments regarding multiple attempts to enter a
solution. Male students stated that when they got a
problem incorrect, they would use subsequent attempts to enter random stuff or manipulations of
an equation, whereas female students stated that
additional attempts gave them an opportunity to
explore multiple approaches to the problem without
stressing out about grades.
4. Course exams and homework
Multiple-choice course exams. Scott, Stelzer, and
Gladding [396] examined the reliability and validity
of using multiple-choice midterm exams in an introductory calculus-based electricity and magnetism

course. They compared students multiple-choice
exam scores to written explanations graded by instructors. The multiple-choice scores were found to be
highly consistent with instructors ratings of the
students explanations, indicating that multiple-choice
exams gave a statistically equivalent assessment of
their understanding (p. 5). Although multiple-choice
exams are easier to grade than free-response questions
or problems, they caution that high-quality multiplechoice questions are more difficult to create. Hudson
and Hudson [395] also found a high correlation
between performance on multiple-choice test questions and free-response questions, suggesting that
multiple-choice tests can provide essentially the
same information as would hand-graded problems
(p. 838). More recent work has begun to explore
whether computer-administered practice exams can
help students better prepare for midterm exams [437].
Homework. A common practice in introductory physics courses is to assign open-ended problems as
homework; however, there are many different approaches for assigning and grading homework (if it is
graded at all). For some courses, especially large
enrollment courses, there is a concern that grading
homework is time consuming and does not provide
timely feedback to students. Bao, Stonebraker, and
Sadaghiani [352] conducted research on a more
flexible homework method in which students have
some freedom to choose which problems to submit,
and half of the problem solutions are available before
the homework is due. Bao et al. [352] found that the
majority of students preferred the new approach and
believed that early access to the solutions improved
their learning. In addition, students who used the
new homework method scored significantly higher
(5%7%) on the Mechanics Baseline Test than students who used traditional homework methods.
One way to provide students with immediate feedback
on their performance might be through the use of electronic
scoring via web-based homework systems. As stated
in the section on Problem Solving, there are several
Web-based homework systems in physics including the
computer-assisted personalized assignment system
CAPA [120,191193], Mastering Physics, WebAssign
[560], and CyberTutor [121]. Studies comparing paperand-pencil homework to Web-based homework have found
mixed results, with some studies citing no differences in
performance [194] and others citing improved performance
with Web-based tutors or homework systems [195198].
Lee et al. [121] used item response theory to analyze
MIT students performance on multiple attempts to answer
a question on the Mastering Physics homework system.
Among students who answered an item incorrectly and
020119-27
received tutoring help, 59% were able to get the solution on

the second attempt, and another 2=3 of students solved it on
their third attempt. Weaker students benefited more from
the tutoringstudents who consulted hints before submitting an answer (even though there was a penalty for this)
showed a large increase of skill, 1.9 standard deviations
better. This study also indicated that the hints provided by
Mastering Physics improved performance on subsequent
attempts of a problem, whereas simply indicating that an
attempt was incorrect was insufficient to improve subsequent performance. In a different study, the same
researchers found a negative correlation between incidence
of homework copying and exam performance [438].
Students who repetitively copied homework performed
worse on exam problems requiring analytic responses.
However, there was no significant difference on normalized
gain for the Mechanics Baseline Test concept inventory.
conditions of the problem, executing mathematical

procedures, and the overall communication or logical
progression of the solution. These skill categories
were found to be consistent with research literature on
problem solving, agree with instructors beliefs, and
are representative of the processes students engage in
while solving physics problems (as measured by
problem-solving interviews). The rubric is applicable
to a range of problem types, physics topics, and
solution formats.
Another effort to assess problem solving includes recent
research by Marx and Cummings [203] and Cummings and
Marx [439] to develop an assessment of students ability to
solve standard textbook-style problems. The draft instrument includes topics of Newtons laws, energy, and
momentum, with 15 items (five from each topic).
Stewart and Ballard [440] assessed students written
presentation on open-response questions on four hourly
exams. The problems were scored along three categories:
math (symbols, relations, and numbers), language (sentences and words), and graphics (i.e., drawings). These
presentation elements could explain 32%26% of variance
in test performance, but could not explain scores on a
conceptual post-test. The incidence of language (sentences
and words) and fewer numbers on a students test was
correlated with better performance.
5. Rubrics for process assessment

A rubric is an assessment tool that indicates a score and
the criteria that must be met to achieve that score. It can be
composed of several subcategories (an analytic rubric) or
can be holistic, and is often formatted as a table or grid.
Rubrics are used in education research to provide criteria
for objective scoring of student responses, but they are also
used in classrooms to communicate expectations to students and help them self-assess their own work.
Assessing scientific abilities. Etkina et al. [272]
developed a set of tasks and rubrics to provide
students with formative feedback on science-process
skills, such as the ability to represent physical processes in multiple ways, design and conduct experimental investigations, and communicate their results.
In classes that used the rubrics, students improved
their abilities to design an experiment, devise a
mathematical procedure to solve an experimental
problem, and communicate procedures. In addition,
these students performed better than students taught
by traditional methods on multiple measures including
use of diagrams in problem solving (60% compared to
20%) and performance on multiple-choice problems
on final tests.
Assessing problem solving. Most measures of problem-solving performance given in the classroom focus
on the correctness of the end result or partial results,
often comparing a students written solution to features of an instructor solution. Scoring criteria and
point values are often determined on a problem-byproblem basis. Docktor [205] developed a more
descriptive problem-solving measure in the form of
a general rubric to evaluate students written solutions
to physics problems along five processes: summarizing problem information and generating a useful
description, selecting physics principles (physics approach), applying physics principles to the specific
6. Complex models of student learning

This section includes efforts to extend score-based
analyses beyond pretestpost-test score comparisons.
Mathematical learning models that consider prior
knowledge. Pritchard, Lee, and Bao [400] present four
mathematical models to describe student learning that
differ in their assumptions about the influence of
students prior knowledge (as measured by pretest
scores). The tabula rasa model assumes that change in
performance (or test gain) is unaffected by prior
knowledge, whereas the constructivist and connectedness models assume prior knowledge affects learning,
and the tutoring model assumes efficient knowledge
acquisition and therefore a high rate of learning. The
researchers found that data from MIT had a lower
value on the connectedness parameter than data
from the University of Minnesota, indicating that MIT
students were using memorization to answer questions
on the FCI.
Model analysis. Bao and Redish [399] propose a
method for analyzing concept inventory scores to
assess the level of confusion in a class. An individual
students responses are characterized mathematically
as a vector, representing the probability they will apply
a particular conceptual model. In addition to indicating
how many students answered a question correctly, this
model analysis approach indicates a level of confusion or the state of a classs knowledge.
020119-28
Learning progressions. A few researchers have investigated the progression of learning throughout an
introductory electricity and magnetism course, looking at the peaks of high scores and decays or
decrease in performance on conceptual questions that
occur between the administrations of pre- and posttests [401,402]. They observed a rapid increase in
performance that occurred after homework assignments (not after lectures or laboratories) but decreased
a few days later. The researchers also observed
interference effects, where learning about electrical
potential impeded students performance on questions
related to the vector nature of electric fields, but
performance recovered later when learning about
magnetic fields.
E. Strengths and limitations of assessment research
1. Strengths
The topic of assessment using multiple-choice inventories lends itself to large-scale, quantitative studies.
Some of the research studies or meta-analyses cited in
this section include test scores for upwards of hundreds or even thousands of students, which provide
confidence in the interpretation of results.
The widespread availability of concept inventories in
physics has facilitated their use in classrooms both
nationally and internationally. Instructors have several
tools they can use to evaluate the effectiveness of their
instruction, and can compare published results from
other institutions to the scores of their students.
From the perspective of making physics instruction
more effective, concept inventories historically have
served a pivotal role in catalyzing dialogs among
physics professors who were initially very skeptical
that students (especially the ones they taught) could
hold erroneous conceptual notions on fundamental
physics ideas following instruction.
2. Limitations
Many studies who report data on inventories do
not give sufficient information about when they
were administered and incentive conditions, which
can have a significant impact on results (see
Ref. [382]).
It is typical for studies to report average scores for a
particular class or group of students, but many do not
indicate the distribution of scores (e.g., histograms) or
separate the results for underrepresented minorities.
Some research on conceptual tests, especially the
Force Concept Inventory, have produced contradictory
results. Although the test authors claim the questions
and choices were developed based on students openended responses to questions, subsequent studies have
found that changing the format of test questions or the
context of the questions can dramatically alter students responses [49,50,427].

Although several sophisticated statistical methods
exist for analyzing test scores and have even been
described in the PER literature (see Ref. [381] for a
review), these procedures have not become widespread, possibly due to a lack of familiarity with
those statistical methods among PER researchers.
Homework and exams. There is very little research on
homework, such as how much to assign, when to
collect it, and how to grade it (if at all). In addition,
studies comparing paper and Web-based homework
have conflicting results, with some stating that Webbased homework is better than paper, and some stating
that they result in equal learning. There is also a need
for research on the type and quality of feedback
provided to students, such as what kind of tutoring
is most helpful in Web-based homework systems.
Similarly, there is a need for research on exams,
including the optimal frequency of administering
quizzes or tests, what types of questions are best
(short answer, essay, multiple choice, problem solving, etc.), and whether comprehensive final exams
have merit. Educational psychology research on the
testing effect suggests that more frequent testing
improves learning and retention of knowledge, but this
has not been studied in science classes [441]. In
addition, there exists little or no research on how
students prepare for physics exams, or the administration and use of practice exams.
Development of additional assessments. As indicated
by the list in Ref. [377], there are several concept
inventories under development and only about half of
the inventories in physics are published. However,
there is some need for alternate forms of assessment
(other than concept inventories), such as tests for
problem solving and mathematics, or measures of
expertise such as categorization tasks. A multiplechoice problem-solving assessment for textbook-style
problems is in development [398,439]. The PER
community also lacks a consistent math skills exam;
several institutions use their own, self-made exam to
assess students skills. Once such measures exist,
additional research could investigate the relationship
between scores on concept tests and problem-solving
performance or the relationship between conceptual
knowledge, mathematical skills, and problem solving.
Gender and other population differences. In general,
there is a need for studies in PER to report data that are
separated for particular populations, such as underrepresented minorities. Research reported in this
section indicates that females score lower on average
than males on some concept inventories like the FCI
020119-29
and FMCE by as much as 10%15%. The reasons

for this gender gap in performance is not well
understood and needs further investigation. For
example, does the gender gap also exist for electricity and magnetism tests (CSEM or BEMA)? Does
it exist for all introductory course populations (nonscience, premedicine, engineers, majors, etc.)? How
are these results related to the concept of stereotype
threat [436]?
Testing conditions. There is a need for additional
research on testing conditions, such as the influence of
when a concept inventory is administered, how it is
administered, and incentives provided to students.
Ding et al. [382] showed for the CSEM test that
even a few days or lectures can influence pretest
scores, and it would be useful to have similar
information for other tests. There have been very
few studies mapping the progression of learning,
decay, and interference that occur during a course
and between pre- and post-tests [401,402]. Also,
the issues involved in administering concept
inventories electronically has only been minimally
explored [426].
Consistent procedures for analyzing and reporting
scores. The normalized gain is a commonly reported measure for comparing pretest and post-test
scores across populations [209], but the statistical
origin for this measure is unclear and alternatives have
been suggested (such as normalized change). It is
unclear why normalized gain is still favored, and
the PER community should reach an agreement
about how to analyze and report scores from concept
inventories.
In addition, the PER community would benefit from
handbooks that outline procedures for the development and
testing of concept inventories, similar to the guidance in
Ref. [378]. These handbooks should also incorporate the
recent interpretations of validity and reliability from the
field of quantitative measurement set forth by [404]
standards for educational and psychological testing
[403]. Many test developers still follow the validity types
paradigm, which is inconsistent with the Standards for
Educational and Psychological Testing.
psychology is usually not important since the focus is on

exploring cognitive processes, and (2) quite often, but not
always, a primary goal in PER studies is the improvement
of teaching and learning in physicsthe improvement of
psychology instruction is almost never the goal in cognitive
psychology research.
Until recently, PER researchers were trained as research
physicists in traditional physics areas and then made the
transition to education research and learned to do PER by
on-the-job experience. Early PER practitioners had little or
no formal training in cognition or education, although
sometimes they were mentored by other physicists interested in or practicing PER. What this means is that
cognitive studies in PER, namely, studies exploring
cognitive processes related to physics learning or problem
solving, have not been common. Although there is some
overlap between cognitive studies, conceptual understanding, and problem solving, articles included in this section
may not have findings that are directly applicable to
improving classroom instruction, yet they carry important
implications for physics learning. Early research on conceptual understanding and problem solving was generally
focused on investigating students difficulties learning
physics and then designing instructional strategies and
curricula to effectively address those difficulties, whereas
cognitive research does not have an explicit focus on
instruction.
The research questions investigated under the cognitive
psychology category generally fall into the following
categories.
1. Knowledge and memory
How is knowledge organized and accessed or activated?
In what ways is knowledge activation influenced by
framing and context? What is the role of specific examples
in knowledge storage and retrieval? Do experts and
novices differ in their organization of knowledge and/or
memory [84,144,151,443,444]?
2. Attention
VI. COGNITIVE PSYCHOLOGY

Cognitive psychology is a field of psychology within
cognitive science [442] focused on studying mental processes such as problem solving, memory, reasoning, learning, attention, perception, and language comprehension. It
emerged about 60 years ago, and in these ensuing decades
the field has matured in terms of theories, methodologies,
and traditions. PER, on the other hand, is not only a
younger field of study, but it is also much more applied than
cognitive psychology in two regards: (1) PER focuses on
physics, whereas the content topic of research in cognitive
Do experts and novices attend to different aspects of

problems [78,141,445]? What do novices attend to when
learning from worked-out problems [143]?
3. Reasoning and problem solving
Do experts and novices approach problem-solving tasks
differently [78]? How do experts and novices categorize
physics problems [84,85]? What is better for developing
problem-solving skills, working problems on ones own or
studying worked-out examples [91]?
020119-30
4. Learning and transfer

What type of knowledge do students transfer and
how does the knowledge transferred depend on context
[446448]? What facilitates and impedes transfer of
learning across physics-similar but surface-different contexts [135,233]? What are possible mechanisms for knowledge transfer, how can they be used to analyze or interpret
interview data, and how does a learner decide that transfer
has been achieved [449,450]? It has been known for some
time that transfer of learning across contexts is difficult to
achieve (see Refs. [451,452] for reviews). Recently a
number of studies have explored various aspects of transfer
in physics contexts.
Given that PER is a relatively young field of study, the
theoretical frameworks used in cognitive studies are borrowed from cognitive science and made to fit the occasion.
Generally, theoretical issues in PER studies analyzing
cognitive processes can be characterized in the following ways:
Expertise and expert performance. The goal of studies
of expertise is to gain insights on effective reasoning
and performance. Although it is recognized that
expertise research can help us think about how to
make the transition from novice to expert more
efficient through the design of instructional interventions, turning novices into experts is not necessarily
the primary goal of this body of research. The study of
expert performance (for reviews, see Refs. [35,453])
has revealed various insights about the way that
experts store and apply knowledge, some of which
will be summarized in Sec. VI D. Research on expert
performance has allowed the characterization of
knowledge in memory for both experts and novices
and how experts and novices use their knowledge to
solve or reason about problems. Studies of expertise
typically compare the performance of experts and
novices on physics tasks such as problem solving or
problem categorization.
Transfer of learning. Transfer studies focus on exploring what knowledge is deployed to reason about a
situation and how the particular knowledge deployed
depends on contextthat is, how and whether or not
changes in context of surface attributes of situations
that do not change the underlying structure affect what
knowledge is brought to bear and/or how it is used in
reasoning. (See Ref. [454] for various studies of
this type.)
Metacognition. Metacognition, which refers broadly
to thinking about ones own thinking [455], is thought
to impact learning in that being reflective about ones
own thinking processes can improve learning and
retention due to deeper processing of the information
[87] (for a broad overview, see Ref. [35]; for a recently
developed assessment of physics metacognition, see
Ref. [456]).
Construction and use of categories. Although it may

seem obvious, proficient problem solving is largely
the ability to categorize new problems into types that
the solver knows how to solve [85,87,153]. There are
several theories regarding the construction and use of
categories in learning, such as the prototype view,
exemplar view, and rule-based views (for a review of
concepts and categories, see Refs. [457,458]).
Attention during processing of information or problem solving. When we look at a situation (e.g., a
picture of diagram) or read a problem, we need to be
selective in what we attend to since short-term
memory is limited (see Ref. [445] for a brief review
and additional references).
Knowledge organization and memory. As we gain
expertise in a subject, that knowledge is organized in
memory in ways that promote quick access or
retrieval. The knowledge organization of experts is
believed to be hierarchical (see Refs. [35,84,151]).
Experts knowledge is bundled with relevant contexts
in which the knowledge can be applied and with
procedures for applying it. With their extensive
experience in problem solving, experts develop problem-type schemas, which consist of representations of
problem categories together with appropriate solution
procedures [144,443,444].
Language. The language people use shapes the way
they think [459]. The language physicists use to
describe and explain physical phenomena is largely
metaphorical. The key here is that metaphors used in
physics discourse were originally analogies with
inherent limitations [460]. As those analogies were
used by physicists, they turned into metaphors that
students perceive as literal [460,461]. For example,
current flows, energy barrier, force acts. In addition,
the grammatical structure of physics statements often
leads to ontological confusion [461,462]. For example, a particle in a potential well, weight of an object,
heat transferred into a system.
Learning by analogy and analogical reasoning.
Studies of use of analogy in making sense of
phenomena and solving problems indicate analogy
is commonly used by peoplethat is, they tend to
look for a similar problem that they know how to solve
or that is already worked out and map how the solution
to the analogous problem can be applied to solve the
new problem. For theories of analogy and analogical
transfer, see Refs. [128,130]. Theories of analogy use
for understanding concepts suggest ways in which
analogical scaffolding can be a useful tool for guiding
students use of representations to understand
phenomena [463465].
C. Methodology, data collection or sources
and data analysis
Various methodologies have been used in physics cognitive studies that are commonly used in cognitive psychology.
020119-31
1. Contexts
Cognitive studies in physics education are difficult to
conduct on a large scale, so they typically take place in a
small controlled environment outside of the classroom.
2. Participants
Participants in cognitive research studies are typically
students who know some physics, often paid volunteers
who have taken one or two introductory courses in physics.
The experts in expert-novice studies are often physics
faculty or graduate teaching assistants who are experienced
with teaching introductory courses.
Research on cognitive processes often requires descriptive, qualitative sources of data, such as interviews or selfexplanations, to infer students reasoning processes as they
engage in a task. Studies on attention might employ
sophisticated data collection tools such as eye-tracking
devices to record gaze patterns. Studies of problem
categorization can utilize a variety of task formats, as
described below.
Verbal reports from clinical interviews. One common
methodology for studying cognitive processes in
physics education research, and many other disciplines as well, is to conduct interviews of subjects
engaged in cognitive tasks, the hope being that verbal
utterances reflect, at least in part, what the mind is
doing [466468]. There are differing views on what
can be ascertained from subjects verbal reports. One
view is that subjects should simply be instructed to
think aloud as they perform a task, since it is
believed that doing so does not change the sequence
of thoughts that would occur if the subject were to
perform the task in silence [453]. Doing more, such as
asking subjects why they make particular choices in
performing a task [469] or asking subjects to selfexplain while reading and attempting to understand a
solution to a problem [470], may change the accuracy
of observed performance in studies of expertise [467].
Others argue that cognitive clinical interviews are
derivative of naturally occurring forms of interaction
and are thus ecologically valid [466], or that the
quality of the information obtained is partly dependent
on students perception of the nature of the discourse
interaction [121]. In PER, interview techniques are
often used that attempt to scaffold or cue students to
previously learned knowledge (see Ref. [450] for
examples and Refs. [5,471]); these interview techniques are useful for learning about ways of structuring effective instructional strategies, as in design
research [472,473].
Self-explanations. Asking subjects to explain to themselves out loud the important features of a problem
solution as they study it has been used to study the

impact of metacognition on learning [87,474].
Eye tracking. Eye tracking methodologies (tracking
eye gazes by capturing or recording position and
duration during cognitive tasks) are quite prevalent in
cognitive psychology and are used to study what
people are paying attention to without disrupting
reasoning processes. There is agreement that where
the eyes look designates what people are paying
attention to, so by following eye fixations one can
unobtrusively observe what is being attended to
during a cognitive task and thereby draw conclusions
about what the mind is processing (see Ref. [475] for a
good review; see also Ref. [476]).
Change events or flicker technique. Methodologies
are beginning to be used in PER that borrow from
those used in visual cognition. For example, change
event methodologies are used to study a phenomenon
known as change blindness. Studies of change blindness [477] in psychology typically involve an unsuspecting subject engaged in a task (e.g., giving
directions to another person, in this case the experimenter, on how to get to a location on a map), and then
unbeknown to the subject something is changed (e.g.,
the experimenter holding a map is changed to another
person holding the same map while the subject is
momentarily distracted) to see if the subject notices.
The flicker technique alternates between two images
on a screen shown for 200 ms with a 100 ms black
screen separating them; the two images are slightly
different and the time to detect the change is measured
to study saliency of the feature being changed. These
methodologies attempt to ascertain what is considered
salient and thus attended to, and what is considered
nonsalient and neglected, or minimally processed, in a
situation or task.
Problem categorization. Problem categorization tasks
are common in physics cognitive studies. These tasks
ask subjects to identify problems that are solved with a
similar approach. Three types of categorization tasks
have been used: (a) problem sorting, where subjects
place into piles problems that are solved similarly
[84,478], (b) problem comparison, where subjects are
shown either two or three problems together and asked
to identify which are solved similarly [85] or are asked
to rate how similar problems are on a Likert scale
[139]. In the two-problem task, subjects are usually
asked to state whether or not the problems are solved
with a similar approach and to state a reason why. In
the three-problem task, a model problem is used and
two comparison problems, and the subject is asked to
identify which of the two comparison problems is
solved most like the model problem; this task allows
experimenting with competing influences on categorization, such as matching a comparison problem to the
020119-32
model problem on either surface attributes, solution

principle or procedure, both, or none. The third
categorization task is (c) multiple choice, where a
problem is given followed by five choices for the
major principle that would be used to solve it and the
task is to select the appropriate one [117].
D. Findings
1. Knowledge and memory
Knowledge activation depends on framing and
context. Framing in anthropology and linguistics
[479482] and more recently science education
[135,483] is used to describe how an individual makes
sense out of context and decides what is important and
what to pay attention to. For example, if framed in
terms of a school context, then students tend to draw
on formal knowledge, whereas if framed in terms of an
out-of-school context, more intuitive or experiential
knowledge is brought to bear [135]; also, nearly
identical tasks can be framed in ways that lead to
very different physics knowledge being brought to
bear [447]. In short, transfer of appropriate knowledge depends highly on framing and context.
Knowledge is tied to specific examples. Since it is
difficult for people to think abstractly, a common
approach in teaching physics and other sciences is to
present the abstract concepts and then follow with a
specific worked example to illustrate how the concepts
are applied to solve problems. In cognitive science,
Ross [131,484] has shown that if following a worked
example students are given a subsequent problem to
solve that is somewhat similar to the worked example,
then the features of the examples context become a
strong cue for students to use the same method or
procedure that they learned in the worked example.
That is, students understanding of the concept becomes bound up in the particulars of the example that
is used to illustrate the principle [132,133,164]. This
phenomenon is known as the specificity effect, and
has been demonstrated in PER [485,486].
Expert-novice differences in knowledge organization
and memory. Experts chunk related knowledge in
clusters, with chunks containing concepts as well as
procedures for applying them and contexts or conditions under which concepts and procedures can be
applied. When recalled, chunks are effective for
reasoning and solving problems since they contain
multiple knowledge, procedural, and contextual units.
In contrast, novices knowledge in memory is more
amorphous and does not have the efficient chunked
organization of experts knowledge; when reasoning
about a physics situation or solving a problem,
novices recall units (e.g., equations, similar contexts)
individually, making the search and thinking processes more hit or miss and simultaneously increasing
memory load by creating high demands on short-term

memory [78,79,81,83,92,93] (for a review, see Chap. 2
in Ref. [35]).
2. Attention
Attending to relevant features in physics diagrams.
Using a change blindness paradigm, Feil and Mestre
[445] investigated whether novices and experts noticed surreptitious changes made to physics diagrams,
finding that experts were able to identify changes that
affected the underlying physics whereas novices did
not. Other studies [141,487] have used eye tracking to
explore how experts and novices allocate visual
attention to physics diagrams in cases where critical
information needed to answer a question was contained in the diagram. They found that experts spend
more time looking at thematically relevant areas of
diagrams than do novices. Other eye-tracking studies
have touched on issues peripherally related to PER,
such as investigations of troubleshooting of malfunctioning circuits [488], of comprehending malfunctioning mechanical devices [489], of comprehending how
mechanical systems work [490], and of the relationship between spatial visualization and kinematics
problem solving ability [491].
Attention in problem solving. Additional eye-tracking
studies using physics tasks are beginning to shed light
on what students, and experts, attend to while performing cognitive tasks. One example is a study by
Rosengrant, Thomson, and Mzoughi [492] in which a
small number of novices and two experts answer
questions about circuit diagrams while their eye gazes
are being recorded. Experts showed a more global
approach to processing information about the circuits
whereas novices focused on local aspects such as
individual resistors to combine in series or parallel.
Another study [143] explored what novices look at
while studying example problem solutions in introductory physics, comparing the time spent on mathematical information (equations) to the textual or
conceptual information. Although students spent a
considerable portion of time (about 40%) looking at
textual or conceptual information, little conceptual
information that students read was retained, suggesting that students may not have been prepared to
learn conceptual knowledge and that this ability may
lag ability to learn mathematical procedures for
solving problems.
3. Reasoning and problem solving
Expert-novice differences in problem solving processes. We refer the reader to the bullet, Problem
solving approaches, within Sec. III D 1 for the salient
differences between expert and novice problem
solving.
020119-33
Problem categorization. As discussed in the

Categorization bullet within Sec. III D 1, experts
categorize physics problems according to the major
concept or principle that can be applied to solve
them, whereas novices rely much more on the
surface attributes of the problems to categorize them
[84,85,145]. Another recent study showed contrasting
results to those of Chi et al. by demonstrating a large
overlap between the categorization performance of
calculus-based introductory students and graduate
students, suggesting that there is a wide distribution
of expertise in mechanics among introductory and
graduate students [493].
Learning from worked examples. In the Example
structure and the worked-example effect bullet in
Sec. III D 2, we discussed the worked-example effect,
namely, research that suggests that studying worked
examples is more effective for learning to solve
problems than actual practice in solving problems.
There is evidence that studying worked examples can
develop schemas better than problem-solving practice
due to the lower levels of cognitive load that go with
studying worked examples, which in turn leaves
more resources in short-term memory to extract and
make sense of solution strategies [494,495]. These
findings have been replicated in other domains,
such as statistics [496] and geometry [497]. The
worked-examples effect holds promise for designing
pedagogical interventions aimed at improving
problem-solving skills in physics.
4. Learning
Self-explanations while studying worked examples.
Research conducted by Chi et al. [87] on how students
study worked-out example solutions in textbooks to
learn topics in mechanics found that successful
students explain and justify solution steps to themselves (self-explain) to a greater extent than poor
students. The quality of the explanations also differs;
good students refer to general principles, concepts, or
procedures which they read in an earlier part of the
text, and examine how they are being instantiated in
the current example [498].
Analogical reasoning. Recent work suggests that use
of analogies during instruction of electromagnetic
waves helps students generate inferences, and that
students taught with the help of analogies outperformed students taught traditionally [463,464].
Further, blending multiple analogies in instruction
generated better student reasoning compared to
instruction that did not use blends or that used
standard abstract representations to convey the wave
concepts [465].
Language and symbolic forms. Research on the role of
language showed that making its metaphorical nature
transparent for the students helps them apply concepts

and solve problems [499]. For example, instead of
writing forces as W (weight) or T (tension), students
benefit when labeling each force with two subscripts
to identify two interacting objectsFEonO (force
exerted by Earth on object) or FRonO (force exerted
by rope on object). Another example is heat. To help
students understand that heat is a process of energy
transfer and of energy itself, the term heat can be
substituted with heating. Some PER-based curriculum materials are using this new language [319].
E. Strengths and limitations of cognitive
research in PER
1. Strengths
PER cognitive psychology research builds upon prior
research from cognitive science and hence has a
corpus of work from which to draw for methodologies
and theoretical frameworks.
This type of research helps us learn about human
cognition in a complex domain (physics) that requires
substantial prior knowledge, reasoning, and problemsolving skills. As such, this type of research can lead
to instructional insights and the eventual design of
effective instructional innovations.
2. Limitations
The findings from many PER cognitive research
studies often are not immediately applicable to improving classroom practice.
Conducting these types of studies most often requires
a pool of subjects who have taken at least one or
several introductory courses since they need to possess some minimal knowledge of physics. This is to be
contrasted with most cognitive psychology experiments that do not require domain knowledge.
Classroom-based cognitive studies involving large
numbers of subjects tend to be rare because it is
much more difficult to control for extraneous variables
that might affect outcomes in these settings; hence, the
majority of cognitive PER studies are done in carefully controlled laboratory settings. This may mean
that results from a lab-based study may yield somewhat different results if one attempts the same study in
a realistic setting at a larger scale.
There are many research areas in cognitive science
whose potential has not been explored for learning about
physics cognition. For example, psycholinguistics is a large
and thriving field of study in cognitive science that draws
heavily upon eye-tracking methodology, yet studies of
020119-34
students understanding of the language of physics is only

now beginning to be explored [461,462]; similar situations
exist with areas such as memory and perception research,
and with promising methodologies, such as electroencephalography (EEG) and functional magnetic resonance
imaging (fMRI), where virtually no PER that we are aware
of exists (the exception being a fMRI study of physics
misconceptions in Ref. [25]). Although subfields (e.g.,
visual cognition) and methodologies (e.g., eye tracking)
from cognitive science are beginning to find their way into
PER, the number of existing studies is extremely small. As
PER matures and broadens its education of Ph.D. candidates to include more training in cognitive science, we will
see an increase in cognitive studies in PER.
beliefs about how students learn physics, instructors

decision-making processes, the ways in which instructors
implement reformed curricula and instruction, and research
on the attitudes and practices of teaching assistants.
VII. ATTITUDES AND BELIEFS ABOUT

TEACHING AND LEARNING, AND THEIR
DIRECT AND INDIRECT IMPACT ON
CURRICULUM CHOICES, CLASSROOM
PRACTICES, AND STUDY HABITS
1. Student attitudes and beliefs about learning physics

What attitudes, beliefs, and expectations do students
have about learning physics? How do students attitudes
and beliefs change as a result of physics instruction? What
instructional strategies are effective for promoting productive attitudes and beliefs? There exists research on
students beliefs about learning physics [42,501,502], and
on the impact that their beliefs have on their performance as
measured by concept inventory scores and course grades
[221,500,503,504]. Research also exists on instructional
strategies designed to explicitly promote productive attitudes and beliefs [224,505]. Below we review several
surveys that have been developed to measure student
attitudes and beliefs before and after instruction, including
the Maryland Physics Expectation Survey (MPEX) [506],
Views About Science Survey (VASS) [503,507],
Epistemological Beliefs about Physical Science (EBAPS)
[224,508], Colorado Learning Attitudes about Science
Survey (CLASS) [509], and Attitudes and Approaches
to Problem Solving survey (AAPS) [510].
Students have attitudes, beliefs, and expectations about

learning physics that can impact the way they behave and
perform in a physics course [221,258,500]. For example, a
common student belief is that physics is made up of several
unrelated pieces of information. As a result, many students
approach physics by memorizing formulas without connecting them to a broader understanding of the underlying
concepts and principles. This section summarizes frameworks that have been developed to describe students
beliefs and multiple-choice attitude surveys to explore
changes that result from instruction.
In addition to students attitudes and beliefs, the beliefs
that instructors (both professors and graduate teaching
assistants) have about how students learn can impact their
instructional decisions and classroom interactions.
Research from PER and science education in general
indicates that oftentimes instructors beliefs and practices
are inconsistent with each other. This section also summarizes research on physics instructors beliefs, their decisions
to adopt research-based curricula, and how PER-based
curricula are actually implemented in the classroom. A
discussion of research on general instructional practices
such as traditional lectures or laboratories has not been
included in this review; there is an abundance of articles on
these topics in the American Journal of Physics dating back
to the 1930s and the Journal of Research in Science and
Teaching dating back to the 1960s. Finally, we include in
this section curricular and instructional decisions by
instructors as well as study decisions by students that,
although perhaps not directly attributable to attitudes and
beliefs, are at least indirectly influenced by them.
How prevalent is the use of research-based instructional

strategies by physics faculty in the United States? How is
[research-based strategy X] typically implemented? What
is the nature of faculty-student interactions during Peer
Instruction? What instructional practices foster an interactive, collaborative classroom environment? Some very
recent studies have explored faculty adoption of PER-based
instructional strategies and materials [368,369,516], how
PER-based strategies like Peer Instruction are actually
implemented in the classroom [517,518], and student
perceptions of classroom reforms.
4. Teaching Assistants
Key areas of research on this topic include student

attitudes and beliefs about learning physics, instructor
What attitudes and beliefs do teaching assistants have

about student learning? How do teaching assistants view
2. Faculty beliefs and values about teaching and learning

How do instructors believe students learn physics (e.g.,
physics problem solving)? What prior knowledge or skills
do instructors expect of their students [511]? How do
instructors view their role as a teacher? What factors
contribute to their instructional decisions [512]? Recent
research has explored beliefs and values that faculty have
about teaching and learning physics in general [512] and
problem solving specifically [201,511,513], including
instructors reasons for choosing problem features [514].
(For a review, see Ref. [515].)
3. Instructor implementations of reformed curricula
020119-35
their role as recitation and/or laboratory instructors? How

do teaching assistants impact the success of reformed
instruction? A few (but growing number of) studies have
explored the attitudes and beliefs that teaching assistants
have about student learning, and how those relate to their
buy-in to reformed instructional strategies and influences
their instructional behaviors [519,520]. Other research
explores how instructional style (specifically TA coaching)
influences student learning during recitation [521].
versus construct knowledge); measures of student success

(against preset standards versus individual improvement);
learning mode (competitive versus cooperative); motivation
(external versus internal); assessments (knowledge based
versus process based); content (facts and principles only
versus also teach thinking and problem solving skills);
instructional design (knowledge driven or student driven);
and problem solving (formulaic versus creative).
and data analysis)
Research studies of attitudes and beliefs about teaching
and learning do not always explicitly identify theoretical
frameworks. Studies might simply cite previous research
findings from education research or psychology as a basis
for their work. Some of the PER research on faculty beliefs
takes this approach (see, for example, Ref. [511]). One
exception is research on teaching assistants [519,520],
which identifies framing theory from anthropology and
linguistics as a basis for understanding how TAs make
sense of their teaching experiences. The following examples are frameworks that have been developed and published by researchers in physics education.
1. Student attitudes and beliefs
Hammer [501,522] developed a framework for characterizing students approaches to learning physics, which
had its basis in education research on learning orientations in addition to interviews he conducted with physics
students. The three categories of his framework are student
beliefs about the structure of physics (made up of pieces
versus coherent), beliefs about physics content (formulas
versus concepts), and beliefs about learning physics
(receiving knowledge versus actively making sense of
new information). Hammers framework was later used
to design multiple-choice attitude surveys.
2. Instructional practices and conceptions
Dancy and Henderson [523] introduce a framework for
describing instructional practices and conceptions. The
framework was developed from research literature, discussions within the PER community, and interviews with
physics faculty. The frameworks ten categories of conceptions include learning view (transmissionist versus
constructivist); beliefs about expertise; beliefs about knowledge; beliefs about the nature of physics; conceptions of
their role as an instructor; the role of the school; views of
students capacity to learn; desired outcomes (goals) for
students; diversity; and scientific literacy. The ten categories of practice identify behaviors typical of traditional
instruction compared to alternative instruction and include
interactivity (passive lecturing versus active student conversation); instructional decisions (made by teacher alone
versus with students); knowledge sources (students receive
1. Contexts
Studies involving students, such as identifying students
attitudes and beliefs about learning physics and the development of attitude surveys, have primarily taken place in
conjunction with introductory physics courses. In contrast,
studies involving instructors (faculty or teaching assistants)
often employ interviews that take place in a controlled
experimental setting outside the classroom or surveys that
are administered online.
2. Participants
Studies of students attitudes and beliefs about learning
physics have typically involved students enrolled in an
introductory course. Studies of instructors beliefs about
learning and teaching physics often involve faculty (from a
variety of institutional types) who are currently teaching a
course or who have substantial experience teaching physics. A very small number of studies involve graduate
teaching assistants in their first or second year of graduate
school.
Clinical interviews. Research to identify the attitudes
and beliefs of students and instructors typically begins
with in-depth interviews that are audio and/or video
recorded. The transcribed statements are coded by
themes that are determined either by the researchers or
from existing theoretical frameworks. These methods
were used by Hammer [47,501,502] to describe
students beliefs about learning physics, and by
several researchers investigating instructors beliefs
[201,511514]. Sometimes these instructor interviews
include specific artifacts such as sample student or
instructor solutions that provide a basis for discussion.
Surveys. Web-based surveys have been used in research on faculty knowledge and use of researchbased instructional strategies [368,369]. The results
are analyzed using descriptive statistical measures,
such as reporting the average fraction of survey
respondents who reported using a particular researchbased instructional strategy.
Classroom observations. Observations of instruction
have been used in studies of teaching assistant
020119-36
behaviors [519,520]. These videotaped observations

are analyzed along several different measures, but
typically attempt to characterize the nature of studentstudent and teacher-student interactions.
Analyzing students written work. In some studies,
students degree of sophistication in their beliefs
about learning physics are analyzed by scoring their
written responses to reflective questions [221,524].
After the researchers have obtained such a quantitative
measure, it is analyzed further by statistical correlations with other course measures such as concept
inventory scores, course exams, homework, or final
course grades.
D. Findings
1. Student attitudes and beliefs about learning physics
This section discusses research on student attitudes and
beliefs and the impact these have on learning physics. In
these studies, a students approach to learning is discussed
using a variety of terms, such as attitudes, beliefs, values,
expectations, views, personal interest, learning orientation,
motivation, and epistemological beliefs. Our use of the
terms attitudes and beliefs is intended to broadly
encompass this wide range in terminology. When the term
epistemology is used, it should be interpreted to refer to
beliefs about what constitutes knowledge in physics and
how, as a student, one develops that knowledge [501].
Some researchers have collectively referred to students
attitudes, beliefs, and assumptions about learning physics
as their expectations upon starting a course [506].
For a more comprehensive review of research on epistemology and relationship to teaching and learning, see
Refs. [525,526].
Epistemological beliefs in physics. Hammer
[42,501,502] developed a framework to characterize
students attitudes and beliefs about physics in
terms of three scales: pieces coherence, formulas
concepts, and by authority independent [205,502].
The coherence measure refers to a students view of
the structure of physics as consisting of several
unrelated pieces of information (unfavorable) or as
a connected system that is relevant to everyday
experiences in the real world (favorable). A students
belief about concepts refers to whether they focus on
memorizing and blindly applying equations (unfavorable) or attempt to understand the concepts that
underlie the equations (favorable). A student with a
favorable attitude on the independence measure
actively attempts to construct their own understanding, whereas a student who just receives information
from an authority (e.g., instructor or textbook) without
question has an unfavorable learning attitude.
Research by Elby [527] extended Hammers ideas to
explore students expectations for a physics course and the
relationship between epistemological beliefs and study
habits. He determined that although most students know

what it means to understand physics deeply and how to go
about learning physics in a deeper way, they do not engage
in productive study habits because they do not see this level
of understanding as essential for obtaining good grades in a
physics class.
Surveys of student attitudes and beliefs. In order to
better characterize student attitudes and beliefs before,
during, or after a course, researchers developed an
assortment of attitude survey instruments. One of the
first surveys developed was the MPEX, a 34-item
Likert-scale questionnaire (disagree-agree) that probes
six dimensions of students expectations about learning physics [506]. The first three dimensions match
those identified by Hammer [501]: independence,
coherence, and concepts; and there are three additional
dimensions: reality link (view physics as relevant to
real life), math link (math is a way to represent
physical phenomena, not just a tool for calculation),
and effort (what students perceive they need to do to
be successful in a physics class). Items are scored as
being favorable if they agree with how an expert
physicist would answer the item, or unfavorable if the
response disagrees with expert responses on that item.
Upon administering the test to physics students at six
universities or schools, they consistently observed a
decline in students attitudes following a semester of
instruction [506]. They suggest that some physics
courses might actually reinforce students inappropriate approaches to learning by placing an emphasis on
memorization rather than a deeper conceptual understanding of the subject matter.
The VASS [503,507] is another survey instrument that
probes students views of learning science along six
dimensions: three scientific dimensions (structure or coherence, methodology or use of mathematics, and validity of
scientific knowledge) and three cognitive dimensions
(learnability or personal effort, reflective thinking, and
personal relevance). They use scores on the VASS to
classify someone as having expertlike views, mixed
(transitional) views, or folk views. For the populations
studied, high school and college students scored similarly
on the survey, and very few (no more than 10%) were
classified as having expertlike views.
The EBAPS is formatted in a slightly different way than
other surveys of attitudes and beliefs [224,508]. Although
part one of the survey includes Likert-style statements
(disagree-agree), part two of the survey includes multiplechoice questions in which students must decide which
statement they agree with most. The five subscales of the
EBAPS include structure of scientific knowledge, nature of
knowing and learning, real-life applicability, evolving
knowledge, and source of ability to learn (natural ability
versus effort based). In contrast to the MPEX, the EBAPS
020119-37
survey attempts to only assess epistemological beliefs

separately from expectations about learning physics.
The most recently developed attitude survey is a
42-item questionnaire adapted from previous work, called
the CLASS [509]. It differs from previous surveys in that
the dimensions are constructed from a factor analysis of
students scores rather than a priori expectations of
question groupings, and it explicitly considers attitudes
toward problem solving. The eight categories include realworld connection, personal interest, sense making or
effort, conceptual connections, applied conceptual understanding, problem solving general, problem-solving confidence, and problem-solving sophistication. There are
also six items that do not fit into these categories and are
not scored. One study of scores on the CLASS [528]
asked students to respond how they think a physicist
would answer the questions, and how they personally
would answer it. When answering from the perspective of
a physicists, students were very accurate at selecting the
favorable attitude responses; however, they did not
answer in the same way when responding with their
own personal beliefs.
Gire, Jones, and Price [529] used the CLASS survey to
measure the epistemological development of physics
majors. They administered the attitude survey to physics
majors who were at different stages in their academic
career, including students in their first, second, third, or
fourth year of college, and graduate students in physics.
They found that among students in their first year of
college, the physics majors had more expertlike
responses on the CLASS than did their introductory
physics course peers, who were primarily composed of
engineering students. Students in years one, two, and three
had similar overall fraction of favorable scores on the
CLASS, whereas the scores were slightly higher for
students in their fourth year or in graduate school.
Another study showed that students at various universities
using the Physics by Inquiry curriculum showed a positive
shift in CLASS scores [530], as did students in a large
introductory course employing active learning [531].
A recently published measure assesses shifts in students
physics course expectations in response to SCALE-UP
orientation and instruction. The assessment is called the
Pedagogical Expectation Violation Assessment (PEVA)
[532]. At the beginning of the course, most students
expected to attend lectures in an amphitheater classroom,
to attend a separate laboratory section, to read the text, and
to memorize equations, all with limited opportunities to
interact with instructors and peers. As a result of a brief
orientation to the course, most students shifted their
expectations to be closer to the actual design of SCALEUP (decreased lecture, an integrated laboratory environment, and more interactions including collaborative group
work). The students also reduced their expectation for
memorizing at two of three institutions studied.
Mason and Singh [510] refined an instrument to measure

attitudes and beliefs about problem solving, called the
AAPS survey. They compared responses on the survey
across several groups: introductory physics and astronomy
students (algebra based and calculus based), graduate
students, and faculty. The graduate students answered
the survey twice, once from the perspective of solving
introductory physics problems, and again from the perspective of solving graduate-level physics problems. In
general, introductory students had beliefs about problem
solving that were less expertlike than the beliefs of graduate
students or faculty. When graduate students took the survey
from the perspective of solving graduate-level physics
problems, their responses were less expertlike than when
they answered the survey for introductory problems. For
example, graduate students indicated that they have difficulty checking whether their answers to graduate problems
are correct, and they feel they need to seek help from others
when they get stuck.
Relationships between epistemological sophistication
and performance. Some studies have explored the
relationship between epistemological beliefs and
understanding of physics concepts as measured by
performance on concept inventories or other course
measures [533]. Halloun [503] reported a significant
correlation between scores on the VASS, FCI gain,
and course grades. For example, students who
were classified as an expert profile on the VASS
survey were most likely to be high achievers in the
physics class.
Perkins et al. [500] observed a statistically significant
correlation between scores on some (but not all) categories
of the CLASS survey and pre- and post-test scores on the
FMCE. The categories of conceptual understanding and
math physics connection were significantly correlated with
FMCE (coefficients in the range of 0.20 to 0.30), whereas
real-world connection, personal interest, and sense making
or effort were not significant (correlation coefficients
ranging from 0.02 to 0.17). In addition, when students
were grouped by normalized learning gains on the FMCE,
students in a high gain bin (>0.9) tended to have more
favorable beliefs whereas students with the lowest gain
(<0.2) had unfavorable beliefs that declined after the
course.
Kortemeyer [524] investigated statistical relationships
between several different measures: epistemology as coded
from students online discussion behavior associated with
Web-based homework and scores on the MPEX survey;
and physics learning as measured by the FCI, final exam,
and course grades. The study found significant correlations
between FCI post-test scores, FCI gain, course grades, and
the extent to which students online discussion posts were
physics oriented (positive) as compared to solution oriented
(negative). These correlations were not observed for the
MPEX, suggesting that online discussions might be a
020119-38
useful (and more authentic) diagnostic tool for assessing

students approaches to learning physics than survey
instruments.
May and Etkina [221] required introductory physics
students to submit weekly reports in which they reflected
on how they learned particular physics topics in electricity
and magnetism. The researchers compared the quality of
students reflections to their performance on several concept inventories: the FCI, MBT, and CSEM. They found
that students with high pre-post normalized gains had a
higher level of epistemological sophistication, such as
asking insightful questions and trying to make sense of
the material rather than saying they learned formulas.
Lising and Elby [534] conducted a case study on a single
student Jan to investigate how epistemology affects
learning. They observed that Jan had a separation between
formal, classroom reasoning and everyday reasoning which
contributed to her difficulties with learning physics.
Although she could reason in both ways, she often did
not attempt to connect them and/or reconcile inconsistencies. They suggest that research-based curricula can be
made even more effective by making epistemology an
explicit part of instruction, especially through reflective
questions on assignments or during class discussions.
Instructional strategies to improve students epistemological beliefs. Elby [224] and Hammer and Elby
[505] developed a set of instructional materials and
strategies to address students epistemological beliefs,
which were found to result in significant, favorable
shifts on the MPEX survey. Effective instructional
practices included assigning essay questions in which
students must argue for or against multiple perspectives, asking students to reflect on whether their
answers agree with their intuition during laboratory
activities, and submitting journal-like paragraphs in
which students reflect about the strategies they use to
learn physics (memorization, summarizing the text,
solving problems, etc.). They acknowledge that implementing these approaches was at the expense of
reduced content coverage, they only minimally used
the textbook, and they attempted to keep lesson plans
flexible to allow time to address students difficulties.
Gender differences on attitude surveys. One study of
scores on CLASS indicate that females have less
expertlike beliefs in statements related to categories of
real-world connections, personal interest, problemsolving confidence, and problem-solving sophistication than their male peers [509]. However, their
responses to questions in the sense-making or effort
category are slightly more expertlike than those of
male students.
teach, but due to external factors these beliefs may not be

reflected in their actual teaching practices [511,512]. An
important first step to understanding instructional decisions
is to identify a common set of beliefs held by instructors
and the situational factors that limit their use of reformed
curricula or teaching methods. This information can then be
taken into consideration by curriculum designers and
leaders of professional development programs when disseminating research-based instructional tools.
Faculty conceptions and instructional practices.
Henderson and Dancy [512] report the results of
semistructured interviews with five experienced, tenured physics faculty from four different institutions.
They analyzed the faculties conceptions about teaching and learning and self-reported instructional practices according to a framework outlined in Dancy and
Henderson [523]. The ten categories used to rate the
conceptions focus on views of student learning and
how the instructor views their teaching role, which
were scored on a scale from being traditional
(transmissionist) in nature to more alternative (constructivist). Similarly, the ten categories used to rate
self-described instructional practices considers
whether an instructor describes their actions as being
more consistent with alternative instruction (e.g.,
active, cooperative, creative, with process-based assessments) or traditional instruction (e.g., passive,
individualistic, formulaic, with knowledge-based assessments). Henderson and Dancy [512] found that
most of the faculty had views about teaching and
learning that were rated as semialternative or a mix
between alternative (reformed) views and traditional
views. In contrast, their descriptions of their own
teaching practices were rated as more traditional in
nature. The study concluded that even though these
faculty members were familiar with research-based
instructional methods, agreed with reform approaches,
and had access to materials, they often did not
implement them because of external factors beyond
their control. These barriers included a need to cover
content (and limited time in which to do so), lack of
time to prepare for teaching, inappropriate class size
and room layout, departmental norms for how
classes are taught, and student factors (student resistance or poor student attitudes).
Beliefs and values about teaching and learning of
problem solving. Problem solving is a primary component to most university physics courses, and a
substantial portion of students grades are often
dependent upon their ability to solve problems as
homework and on exams. Students are provided with
worked-out examples in their textbooks and might
observe additional demonstrations of problem solving
during class; however, the problem features, example
solutions, and problem-solving procedures used by
2. Faculty beliefs and values about teaching and learning

Faculty beliefs and values about how students learn
physics influence their decisions about what and how to
020119-39
different instructors may vary widely. Yerushalmi

et al. [511] and Henderson et al. [513] used structured
interviews with six physics faculty to explore factors
that contributed to their instructional decisions related
to the teaching and learning of problem solving. The
interviews took place with faculty at a research
university and utilized specific artifacts as a basis
for discussion, including different formats for the
same problem statement, multiple student solutions,
and multiple versions of an instructor solution [513].
During the interviews, the participants were prompted
to comment on their format preferences for each type
of artifact and indicate what they typically use in the
classes they teach and why.
The researchers concluded that these physics instructors
had unstable, often conflicting beliefs that were constructivist in nature, while their actions in the classroom
reflected a traditional model of transmitting information
[511]. For example, the faculty expressed a belief that
students must be reflective learners and solve a lot of
problems on their own to gradually build up an understanding of physics (an inquiry-based, constructivist view),
but they often provided explicit guidance in course materials such as using problems that are broken into parts (parts
a, b, c, etc.) to direct students through a problem-solving
procedure. In addition, the faculty experienced a conflict
between their physicist view that places value on
compact, concise problem solutions and a teacher value
of wanting students to communicate their reasoning. As a
result, the instructors were unwilling to penalize a student
who wrote a very sparse answer that could be interpreted as
a correct result, but would penalize students who wrote
incorrect reasoning that led to a correct numerical result
[201]. The faculty who were interviewed acknowledged
that their preferences do not match what they use in their
courses, often because it takes too much time and expertise
to construct high-quality problems and problem solutions,
and they also do not want to overwhelm students with
complicated or overly detailed problem solutions.
A later study extended this research to include interviews
with 30 physics faculty at a wide range of institutions, and
focused on instructors choice of features when writing or
selecting physics problems for their class [514]. During
interviews, faculty indicated their preferences for or against
particular problem features, including problems that were
qualitative (no calculation), multiple choice, broken into
parts, had a real-world context, were wordy (extra unnecessary words), included a drawing, or were complex
(required multiple principles for a solution). The interview
statements were coded along three dimensions: whether a
particular problem feature supported or hindered an instructors teaching goals, whether the feature was used by a
faculty member in their courses in any context, and whether
the feature was used on exams. The researchers concluded
that for four of these features, the instructors values were in
conflict with their practices. Although faculty believed

conceptual questions promote student understanding and a
real-world context provides necessary motivation for students, they chose not to use problems with those features in
class. In contrast, they stated that problems that are broken
into parts and include a drawing hinder their goal that
students will learn general problem-solving skills, but they
frequently use these features to make a problem clear,
especially on exams.
3. Instructor implementations of reformed curricula
This section reviews findings from studies related to
faculty adoption of PER-based instructional strategies and
materials, how PER-based strategies are actually implemented in the classroom, and student perceptions of
classroom reforms.
Self-reported knowledge and use of reformed instruction. Henderson and Dancy [368] conducted a largescale online survey of physics faculty in the U.S. Of
the 722 faculty who responded, most of them had
heard of at least one research-based instructional
strategy (87.1%) and about half were familiar with
six or more strategies. In terms of use, approximately
half of the faculty reported using reformed instruction
in their teaching, but often with significant modifications. From the list of 24 research-based instructional
strategies, the most common instructional strategy was
peer instruction, with 63.5% of faculty reporting they
know about the strategy and 29.2% of them saying
they use it in some form. Other strategies that were
familiar to more than 40% of faculty include Physlets,
cooperative group problem solving, workshop physics, just-in-time teaching, tutorials, and interactive
lecture demonstrations. Although this survey shows
that there is high knowledge of PER-based instruction
among faculty, situational factors such as limited time
or lack of support deter faculty from using them
[368,512]. During interviews with five physics faculty, Henderson and Dancy [516] also found that
several faculty expressed dissatisfaction with their
interactions with educational researchers. They suggest that PER should change its model for disseminating information and materials to better help faculty
adapt the strategies to their specific situation.
Implementations of Peer Instruction. Turpen and
Finkelstein [517] used detailed classroom observations to identify variations in the ways that physics
instructors implement Peer Instruction [213]. They
found that the basic elements of Peer Instruction were
present in all six classes studied: the presentation of
conceptual information (usually in the form of questions) and opportunities for students to discuss physics
with their classmates. The instructors differed, however, in the average number of conceptual questions
asked per class period (range from 3 to 8), how they
020119-40
interacted with students (both during the voting time

and discussing the responses), and the time spent
explaining the solution (range from 1 minute to nearly
4 minutes). One of the most prominent differences
was in the average number of explanations heard from
students during a class period, which was 4 or 5
statements for three of the instructors, around 2 or 3
for one instructor, and between 0 and 1 for the
remaining two instructors. In a related study, Turpen
and Finkelstein [518] administered surveys to students
to explore the relationship between these instructor
differences and student perceptions of their classroom
experiences. They found that students who were in a
class with more opportunities for discussion felt
comfortable asking questions and speaking during
class, with both their peers and the instructor. A higher
fraction of these students also stated that it was
important to understand the reasoning for an answer,
not just to know the correct answer.
classes taught by teaching assistants using one unit

from the Tutorials in Introductory Physics curriculum
and four different versions of implementation. The
topic studied was energy and momentum, and the four
instructional modes were lecture presentation of the
materials, students working individually on the tutorial, students working in a group of 3 or 4 students,
and students working in a group with instructor
coaching (Socratic dialog). Student learning in each
of these classes was assessed using pretestpost-test
questions designed by the authors, which were written
to assess concepts addressed during the tutorial unit.
They found that students who experienced the fourth
style of instruction scored significantly higher on the
post-test than other groups (cooperative groups with
instructor coaching). Surprisingly, the students in the
third style (cooperative learning without instructor
dialogue) scored at the same level as students in the
lecture or individual course sections, suggesting that
the TA interactions with students had a substantial
influence on learning.
4. Teaching assistants
Graduate and undergraduate students are frequently
employed to teach laboratory and/or recitation sections
of introductory physics courses, especially at large institutions. The preparation and instructional support these
TAs receive varies widely across institutions, and to date
limited research has been done on teaching assistants
beliefs and practices.
Teaching assistants pedagogical beliefs and their
influence on instructor-student interactions. Goertzen,
Scherr, and Elby [519,520] and Spike and Finkelstein
[535] used interviews and videotaped observations
of teaching assistants to analyze their classroom
interactions while using Tutorials curriculum at the
University of ColoradoBoulder (CU) and the
University of Maryland (UM). Goertzen et al. found
that a lack of buy-in from the TA resulted in
behaviors that conflicted with the tutorial developers
intentions, such as giving direct answers to students
questions and increasing the use of equations as a
reasoning tool. Spike and Finkelstein concluded that
the beliefs and behaviors of teaching assistants had a
profound impact on the attitudes of students in the
class, such as their willingness to engage in group
work. Additional studies have investigated the pedagogical beliefs of graduate students and learning
assistants about problem solving [536], studio style
classrooms [537], interactive-engagement courses
[538], or an inquiry-based course for future elementary school teachers [539]. These studies have generally concluded that there is a wide range in teaching
assistant behaviors, which result in a variety of
different experiences for students in those classes.
The impact of instructional styles on student learning.
Koenig, Endorf, and Braun [521] studied recitation
E. Strengths and limitations of research

on attitudes and beliefs
1. Strengths
There has been at least one large-scale survey research
study to identify what fraction of faculty know about
PER-based instructional strategies and how many
claim to use them [368,369]. This gives a baseline
indication of the success attained by dissemination
efforts in PER and where future efforts should be
focused. This, as well as interview studies of adoption
of PER-informed instructional strategies, provides
insights into what is needed to accelerate the adoption
of effective curricula in teaching undergraduate
physics.
The prevalence of student attitude surveys (like the
MPEX and CLASS) has given instructors an easy way
to monitor students attitudes before, during, and after
physics instruction.
2. Limitations
Qualitative studies involving video- and audiotaped
interviews or observations are time consuming and
difficult to analyze, so there is often a delay in the
availability of results from these studies.
Some attitude surveys, like the VASS and EBAPS, do
not have published documentation about the instruments development and score validation process.
Research into faculty adoption of reformed curricula
is in its infancy stage and a lot more work is needed to
understand ways of helping faculty change their
attitudes and adopt more effective instructional
strategies, as well as ways of helping departments
020119-41
implement reward structures for implementing evidence-based reforms effectively.

Varied implementations of reformed curricula.
Although this section summarizes research studies
that describe instructors variations in their implementation of certain interactive-engagement strategies
(such as Peer Instruction), there is a need for studies
that explore how instructors modify and implement
other research-based instructional strategies. There is
also research needed on sociological and infrastructure factors that interfere with adoption of reformed
curricula and instructional practices for faculty, undergraduate learning or teaching assistants, and graduate
teaching assistants.
Teaching assistants. There is a need for additional
research on the attitudes and beliefs of TAs, how
physics TAs impact the success of instructional reforms, and the implications for professional development of TAs. There is little guidance on how to
prepare teaching assistants, in terms of both the
components of a teaching assistant orientation program and ongoing professional development opportunities during teaching.
Institutional barriers. There is a need for research on
approaches that research universities can take to
encourage and reward the implementation of researchbased instructional strategies. Advancement in
research universities for faculty (e.g., achieving tenure
and promotion, receiving merit raises, being respected
among peers) largely hinges on professors research
prowess, not teaching prowess. Better instruction
would likely result when reward structures align
with teaching accomplishments that reflect effective
student learning, but it is not clear how to change
institutional structures and traditions in ways that are
acceptable to administrators, faculty, and institutional
goals and values.
VIII. SUMMARY AND CONCLUSIONS
1. Summary of research in physics education
from an historical perspective
Even with this nonexhaustive review of PER at the
undergraduate level, it is evident that there has been a
considerable corpus of research done to date. One characterization of PER is that it had its origin in some instructors
observing stubborn conceptual difficulties experienced by
their students in topics that seemed to the instructors to be
simple physics ideas. This interest in digging deeper into
students conceptual learning difficulties led to hundreds of
studies on common misconceptions, which have been
documented for nearly every physics topic ranging from
mechanics concepts like force, motion, momentum, and
energy to topics in electricity and magnetism, thermal

physics, light and optics, and modern physics. Research to
document students conceptual difficulties in the 1970s and
1980s led to the development of multiple-choice concept
inventories that became widely available in the early 1990s,
and to instructional interventions aimed at helping students
overcome conceptual difficulties.
The availability of concept inventories served as a
catalyst for discussions among physicists about the nature
of learning and of conceptual difficulties among their
students. Developers of early concept inventories, as well
as PER researchers, invited colleagues to administer those
inventories to their students following instruction; most
refused, considering it to be a waste of time since they
thought that surely their students would not hold these
incorrect notions about fundamental and easy (to the
professors) physics ideasafter all, they were good
instructors who presented ideas clearly. Those few who
initially took up the offer were surprised, perhaps even
dismayed, by what they found. What later research showed
was that quality of lecturing or instructor charisma had little
to do with helping students learn concepts about which they
held deeply rooted beliefs that contradicted physics laws.
Parallel with research into conceptual difficulties was
interest in problem solving, given how central it is to
physics. Initial research studies in the late 1970s and early
1980s focused on describing expert-novice differences in
problem solving, by contrasting the processes used by
beginning physics students to the principle-based
approaches used by experienced solvers. These studies
led to the development of instructional strategies and
curricula to promote the use of expertlike approaches,
which continue to be active topics of research today. It has
also not gone unnoticed with physics instructors that
teaching problem-solving skills to students in physics is
a challenging endeavor. Indeed, nearly all physics instructors have experienced students coming to them and stating
that they are A students in nearly all subjects but that they
are doing poorly in physics and pleading for some
prescriptive guidance for how to do well on exams.
Thus, research on problem solving combined with
research on conceptual understanding has given rise to
research-based and research-informed concept inventories,
curricula, and instructional strategies. Research-based
instructional strategies have become collectively referred
to as interactive engagement methods in contrast to
traditional, passive modes of instruction. For example,
the use of classroom polling technologies or clickers
and interactive demonstrations have become relatively
widespread in introductory physics courses, and classroom
environments structured to support increased interactions
are growing in popularity (e.g., workshop, studio, and
SCALE-UP classrooms). Several curricular packages
were identified in Sec. IV, and some of these have been
in existence for several decades. In recent years the design
020119-42
of curricula has expanded to include computer-based and

Web-based instruction, including online homework systems, animations or simulations of phenomena, multimedia
presentations, and computer tools for laboratory data
collection and analysis.
As some of the more obvious things to try have already
been done in PER, the decade from 2000 to 2010 has seen
increased work in more interdisciplinary areas, such as
applications of cognitive psychology to physics learning,
argumentation in physics from a linguistic or psycholinguistic perspective, and student or faculty attitudes about
teaching and learning, including opportunities and
obstacles for instructors adoption of research-based or
informed instructional strategies.
Besides looking at the expansion of PER, another way to
characterize the growth and interest in PER is to observe the
historical development of the Physics Education Research
Conference. Prior to 2001, PER conferences were a
cottage industry with sporadic conferences held whenever a senior member of the PER community decided to
organize one. There was one such conference organized by
Beichner at NC State University in 1994 attracting about 43
attendees (24 PER faculty, with the remaining19 being
graduate students and local faculty); the venue was not to
discuss research, but issues such as the job market, what
graduates of PER programs should know, how they should
learn it, what the requirements of a PER graduate program
should be, the need for publication venues, future conferences, research funding, and educating the physics community. The next PER conference was organized by Fuller
at the University of Nebraska in 1998 attracting 83
participants. Starting in 2001, the Physics Education
Research Conference (PERC) became a yearly event
attached to the end of the American Association of
Physics Teachers summer meeting; a peer-reviewed
conference proceeding also became an ancillary part of
the conference. Typical attendance at the PERC during the
last few years has been over 200.
of todays textbooks; in contrast, other than the addition of

color and other cosmetic changes, introductory physics
textbooks have changed little. Physics instruction also
places a strong emphasis on quantitative problem solving,
which is not the case in other science disciplines (e.g.,
biology and introductory astronomy). Engineering places a
major emphasis on design, while field experiences and a
systems approach are very prominent in geology; not so in
traditional physics instruction. These differences are likely
to have important implications for the type and direction of
STEM discipline-based education research.
In summary, we believe PER has a strong research basis
in the first four areas addressed in this review: students
difficulties learning physics concepts and solving problems, the development and evaluation of instructional
strategies and curricula (particularly interactive engagement methods), and the design and analysis of concept
inventories. The final two areas of cognitive psychology
and attitudes and beliefs are less developed but growing.
The Supplemental Material [3] for this synthesis includes
suggestions and speculations about future directions for
PER that could hold promise, or that at least would further
inform physics teaching and learning.
In conclusion, physics education research as a discipline
has a rich history, makes a wide array of interdisciplinary
connections, and yet has many promising avenues for
future research.
2. What distinguishes PER from other DBER Fields?

For us, it is hard to characterize PER as one thing since it
is diverse and evolving, drawing from disciplines such as
cognitive science, sociology, education, linguistics,
psycholinguistics, assessment, and measurement.
It should be noted that there are some elements of
physics that distinguish it from other natural sciences,
which may have implications for how PER differs from
other discipline-based educational research. One is that the
content of introductory physics (i.e., classical physics) has
changed little in more than a century. This is certainly not
the case in the biological sciences, astronomy, and geology,
where the content of textbooks from 50 years ago has
undergone major transformations compared to the content
ACKNOWLEDGMENTS
We thank the National Research Council, and in particular Natalie Nielsen and Susan Singer from the NRCs
Committee on the Status, Contributions and Future
Directions of Discipline-Based Education Research, for
asking us to develop the original PER synthesis from which
this publication emerged and for their very helpful discussions that helped shape the structure of this synthesis.
We also thank Charles Henderson for encouraging us to
update the original NRC white paper for publication in
PRST-PER. Additionally, we thank the anonymous
reviewers for their helpful suggestions. We also thank
Paula Heron, Michael Wittmann, and Rachel Scherr for
inviting us to present a summary of our white paper at the
2011 Frontiers and Foundations of Physics Education
Research Conference, and Suzanne Brahmia for helping
to summarize the suggestions made by attendees at this
conference; we found the interactions and suggestions
made by the attendees of this conference extremely helpful.
Finally, we thank Jeanne Brunner, for her help in updating
our review of the literature. This material is based on work
supported by the National Science Foundation under Grant
No. DUE-1347722.
020119-43
[1] J. Docktor and J. Mestre, A Synthesis of Discipline-based

Education Research in Physics (National Academy Press,
Washington, DC, 2011), http://sites.nationalacademies
.org/dbasse/bose/dbasse_080124.
[2] S. R. Singer, N. R. Nielsen, and H. A. Schweingruber,
Discipline-based Education Research: Understanding
and Improving Learning in Undergraduate Science
and Engineering (National Academy Press, Washington
DC, 2012).
[3] See Supplemental Material at http://link.aps.org/
supplemental/10.1103/PhysRevSTPER.10.020119
for
the authors subjective views of promising future areas
for PER research.
[4] J. J. Clement, Students preconceptions in introductory
mechanics, Am. J. Phys. 50, 66 (1982).
[5] L. C. McDermott, Research on conceptual understanding
in mechanics, Phys. Today 37, 24 (1984).
[6] Proceedings of the International Seminar on
Misconceptions in Science and Mathematics, edited by
H. Helm and J. D. Novak (Department of Education,
Cornell University, Ithaca, NY, 1983).
[7] Proceedings of the Second International Seminar on
Misconceptions and Educational Strategies in Science
and Mathematics, Vols. I, II, and III, edited by J. D.
Novak (Department of Education, Cornell University,
Ithaca, NY, 1987).
[8] S. Flores-Garcia, L. L. Alfaro-Avena, J. E. Chavez-Pierce,
J. Luna-Gonzalez, and M. D. Gonzalez-Quezada,
Students difficulties with tension in massless strings,
Am. J. Phys. 78, 1412 (2010).
[9] R. A. Streveler, T. A. Litzinger, R. L. Miller, and P. S.
Steif, Learning conceptual knowledge in the engineering
sciences: Overview and future research directions, J. Eng.
Educ. 97, 279 (2008).
[10] L. C. McDermott and E. F. Redish, Resource letter:
PER-1: Physics education research, Am. J. Phys. 67,
755 (1999).
[11] B. S. Ambrose, P. R. L. Heron, S. Vokos, and L. C.
McDermott, Student understanding of light as an electromagnetic wave: Relating the formalism to physical
phenomena, Am. J. Phys. 67, 891 (1999).
[12] L. C. McDermott and P. S. Shaffer, Research as a guide
for curriculum development: An example from introductory electricity. Part I: Investigation of student understanding, Am. J. Phys. 60, 994 (1992).
[13] R. J. Beichner, Testing student interpretation of kinematics graphs, Am. J. Phys. 62, 750 (1994).
[14] D. Hestenes and M. Wells, A mechanics baseline test,
Phys. Teach. 30, 159 (1992).
[15] D. Hestenes, M. Wells, and G. Swackhamer, Force
Concept Inventory, Phys. Teach. 30, 141 (1992).
[16] D. P. Maloney, T. L. OKuma, C. J. Hieggelke, and
A. Van Heuvelen, Surveying students conceptual knowledge of electricity and magnetism, Am. J. Phys. 69, S12
(2001).
[17] R. K. Thornton and D. R. Sokoloff, Assessing student
learning of Newtons laws: The Force and Motion
Conceptual Evaluation and the Evaluation of Active
Learning Laboratory and Lecture Curricula, Am. J. Phys.
66, 338 (1998).
[18] M. T. H. Chi and J. D. Slotta, The ontological coherence

of intuitive physics, Cognit. Instr. 10, 249 (1993).
[19] M. T. H. Chi, J. D. Slotta, and N. de Leeuw, From things
to processes: A theory of conceptual change for learning
science concepts, Learn. Instr. 4, 27 (1994).
[20] A. A. diSessa, N. Gillespie, and J. Esterly, Coherence vs.
fragmentation in the development of the concept of force,
Cogn. Sci. 28, 843 (2004).
[21] A. Gupta, D. Hammer, and E. F. Redish, The case for
dynamic models of learners ontologies in physics,
J. Learn. Sci. 19, 285 (2010).
[22] C. Ioannides and S. Vosniadou, The changing meanings
of force, Cognit. Sci. Q. 2, 5 (2002).
[23] J. D. Slotta, M. T. H. Chi, and E. Joram, Assessing
students misclassifications of physics concepts: An
ontological basis for conceptual change, Cognit. Instr.
13, 373 (1995).
[24] G. Chiou and O. R. Anderson, A study of undergraduate
physics students understanding of heat conduction based
on mental model theory and an ontology-process analysis,
Sci. Educ. 94, 825 (2010).
[25] K. N. Dunbar, J. A. Fugeslsang, and C. Stein, in Thinking
with Data, edited by M. C. Lovett and P. Shah (Erlbaum,
Mahwah, NJ, 2007), pp. 197205.
[26] C. Camp and J. Clement, Preconceptions in Mechanics:
Lessons Dealing with Students Conceptual Difficulties
(Kendall-Hunt, Dubuque, IA, 1994).
[27] L. C. McDermott, P. S. Shaffer, and The Physics
Education Group at the University of Washington,
Tutorials in Introductory Physics (Prentice-Hall, Upper
Saddle River, NJ, 2002).
[28] J. Slotta and M. T. H. Chi, Helping students understand
challenging topics in science through ontology training,
Cognit. Instr. 24, 261 (2006).
[29] D. R. Sokoloff and R. K. Thornton, in Proceedings of
ICUPE, edited by E. F. Redish and J. S. Rigden
(American Institute of Physics, Melville, NY, 1997),
pp. 10611074.
[30] M. S. Steinberg and C. L. Wainwright, Using models to
teach electricityThe CASTLE project, Phys. Teach. 31,
353 (1993).
[31] K. Wosilait, P. R. L. Heron, P. S. Shaffer, and L. C.
McDermott, Development and assessment of a
research-based tutorial on light and shadow, Am. J. Phys.
66, 906 (1998).
[32] E. D. Corpuz and N. S. Rebello, Investigating students
mental models and knowledge construction of microscopic friction. II. Implications for curriculum design and
development, Phys. Rev. ST Phys. Educ. Res. 7, 020103
(2011).
[33] G. Posner, K. Strike, P. Hewson, and W. Gerzog,
Accommodation of a scientific conception: Toward a
theory of conceptual change, Sci. Educ. 66, 211 (1982).
[34] E. Etkina, J. Mestre, and A. ODonnell, in The Cognitive
Revolution in Educational Psychology, edited by J. M.
Royer (Information Age Publishing, Greenwich, CT,
2005), pp. 119164.
[35] J. D. Bransford, A. L. Brown, and R. R. Cocking, How
People Learn: Brain, Mind, Experience, and School
(National Academy Press, Washington, DC, 1999).
020119-44
[36] S. Carey, in Conceptual Development: Piagets Legacy,

edited by E. K. Scholnick, K. Nelson, and P. Miller
(Lawrence Erlbaum Assoc., Mahwah, NJ, 1999),
pp. 293326.
[37] G. M. Sinatra and P. R. Pintrich, in Intentional
Conceptual Change, edited by G. M. Sinatra and
P. R. Pintrich (Lawrence Erlbaum Assoc., Mahwah, NJ,
2003), pp. 118.
[38] K. A. Strike and G. J. Posner, Conceptual change and
science teaching, Eur. J. Sci. Educ. 4, 231 (1982).
[39] K. A. Strike and G. J. Posner, in Philosophy of Science,
Cognitive Psychology, and Educational Theory and
Practice, edited by R. A. Duschl and R. J. Hamilton
(State University of New York Press, New York,
1992), pp. 147176.
[40] A. A. diSessa, Toward an epistemology of physics,
Cognit. Instr. 10, 105 (1993).
[41] A. A. diSessa and B. L. Sherin, What changes in conceptual change? Int. J. Sci. Educ. 20, 1155 (1998).
[42] D. Hammer, Misconceptions or p-prims: How might
alternative perspectives on cognitive structure influence
instructional perceptions and intentions, J. Learn. Sci. 5,
97 (1996).
[43] Harvard-Smithsonian Center for Astrophysics, A Private
Universe, videotape, http://www.learner.org/resources/
series28.html, retrieved Sept. 15, 2010.
[44] D. Hammer, Student resources for learning introductory
physics, Am. J. Phys. 68, S52 (2000).
[45] J. Minstrell, in Proceedings of the International
Workshop on Research in Physics Education: Theoretical
Issues and Empirical Studies, edited by R. Duit,
F. Goldberg, and H. Niedderer (Institut fur die
Padagogik der Naturwissenschaften, Kiel, Germany,
1992), pp. 110128.
[46] J. Smith, A. diSessa, and J. Roschelle, Misconceptions
reconceived: A constructivist analysis of knowledge in
transition, J. Learn. Sci. 3, 115 (1994).
[47] D. Hammer, More than misconceptions: Multiple perspectives on student knowledge and reasoning, and an
appropriate role for education research, Am. J. Phys. 64,
1316 (1996).
[48] D. Hammer and A. Elby, in Personal Epistemology: The
Psychology of Beliefs about Knowledge and Knowing,
edited by B. K. Hofer and P. R. Pintrich (Erlbaum,
Mahwah, NJ, 2002), pp. 169190.
[49] R. J. Dufresne, W. J. Leonard, and W. J. Gerace, Making
sense of students answers to multiple choice questions,
Phys. Teach. 40, 174 (2002).
[50] R. N. Steinberg and M. S. Sabella, Performance on
multiple-choice diagnostics and complementary exam
problems, Phys. Teach. 35, 150 (1997).
[51] D. E. Brown and D. Hammer, in International Handbook of
Research on Conceptual Change, edited by S. Vosniadou
(Routledge, New York, 2008), pp. 127154.
[52] M. T. H. Chi, in Cognitive Models of Science: Minnesota
Studies in the Philosophy of Science, edited by R. Giere
(University of Minnesota Press, Minneapolis, MN, 1992),
pp. 129186.
[53] M. T. H. Chi, in Conceptual Structures, and Processes:
Emergence, Discovery, and Change, edited by T. B. War,
[54]
[55]
[56]
[57]
[58]
[59]
[60]
[61]
[62]
[63]
[64]
[65]
[66]
[67]
[68]
[69]
[70]
[71]
020119-45
S. M. Smith, and J. Vaid (American Psychological

Association, Washington, DC, 1997), pp. 209234.
F. C. Keil, Semantic and Conceptual Development: An
Ontological Perspective (Harvard University Press,
Cambridge, MA, 1979).
G. Zhu and C. Singh, Improving students understanding
of quantum measurement. I. Investigation of difficulties,
Phys. Rev. ST Phys. Educ. Res. 8, 010117 (2012).
G. Zhu and C. Singh, Improving students understanding
of quantum measurement. II. Development of researchbased learning tools, Phys. Rev. ST Phys. Educ. Res. 8,
010118 (2012).
A. Strauss and J. Corbin, Basics of Qualitative Research
(Sage, Thousand Oaks, CA, 1998).
D. Brown and J. Clement, Overcoming misconceptions
via analogical reasoning: Abstract transfer versus
explanatory model construction, Instr. Sci. 18, 237
(1989).
J. Clement, Using bridging analogies and anchoring
intuitions to deal with students preconceptions in
physics, J. Res. Sci. Teach. 30, 1241 (1993).
Harvard-Smithsonian Center for Astrophysics, Minds of
Our Own: Lessons from Thin Air, videotape, http://www
.learner.org/resources/series26.html, retrieved Sept. 15,
2010.
M. T. H. Chi, Common sense conceptions of emergent
processes: Why some misconceptions are robust, J. Learn.
Sci. 14, 161 (2005).
M. Chiu, C. Chou, and C. Liu, Dynamic processes of
conceptual changes: Analysis of constructing mental
models of chemical equilibrium, J. Res. Sci. Teach. 39,
688 (2002).
D. B. Clark, Longitudinal conceptual change in students
understanding of thermal equilibrium: An examination of
the process of conceptual restructuring, Cognit. Instr. 24,
467 (2006).
M. Ferrari and M. T. H. Chi, The nature of nave
explanations of natural selection, Int. J. Sci. Educ. 20,
1231 (1998).
C. Furio and J. Guisasola, Difficulties in learning the
concept of electric field, Sci. Educ. 82, 511 (1998).
E. Kikas, Teachers conceptions and misconceptions
concerning three natural phenomena, J. Res. Sci. Teach.
41, 432 (2004).
Y. Lee and N. Law, Explorations in promoting conceptual
change in electrical concepts via ontological category
shift, Int. J. Sci. Educ. 23, 111 (2001).
L. M. Tyson, G. J. Venville, A. G. Harrison, and D. F.
Treagust, A multidimensional framework for interpreting
conceptual change events in the classroom, Sci. Educ. 81,
387 (1997).
G. J. Venville and D. F. Treagust, Exploring conceptual
change in genetics using a multidimensional interpretive
framework, J. Res. Sci. Teach. 35, 1031 (1998).
L. Deslauriers and C. Wieman, Learning and retention of
quantum concepts with different teaching methods, Phys.
Rev. ST Phys. Educ. Res. 7, 010101 (2011).
F. Reif and L. Scott, Teaching scientific thinking skills:
Students and computers coaching each other, Am. J.
Phys. 67, 819 (1999).
[72] F. Reif and M. St. John, Teaching physicists thinking

skills in the laboratory, Am. J. Phys. 47, 950 (1979).
[73] A. Van Heuvelen, Learning to think like a physicist: A
review of research-based strategies, Am. J. Phys. 59, 891
(1991).
[74] D. P. Maloney, in Handbook of Research on Science
Teaching and Learning, edited by D. L. Gabel
(MacMillan, New York, 1994), pp. 327354.
[75] L. Hsu, E. Brewe, T. M. Foster, and K. A. Harper,
Resource letter RPS-1: Research in problem solving,
Am. J. Phys. 72, 1147 (2004).
[76] K. Cummings, A Developmental History of Physics Education Research (National Academy Press, Washington,
DC, 2011), http://sites.nationalacademies.org/dbasse/bose/
dbasse_080124.
[77] E. Bagno and B. Eylon, From problem solving to
knowledge structure: An example from the domain of
electromagnetism, Am. J. Phys. 65, 726 (1997).
[78] M. T. H. Chi, R. Glaser, and E. Rees, in Advances in the
Psychology of Human Intelligence, edited by R. J.
Sternberg (Lawrence Erlbaum Associates, Hillsdale,
NJ, 1982), Vol. 1, pp. 775.
[79] B. Eylon and F. Reif, Effects of knowledge organization
on task performance, Cognit. Instr. 1, 5 (1984).
[80] J. Larkin, Processing information for effective problem
solving, Eng. Educ. 70, 285 (1979).
[81] J. Larkin, Cognition of learning physics, Am. J. Phys. 49,
534 (1981).
[82] J. Larkin, in Cognitive Skills and Their Acquisition,
edited by J. R. Anderson (Lawrence Erlbaum Associates,
Hillsdale, NJ, 1981), pp. 311344.
[83] J. H. Larkin, J. McDermott, D. P. Simon, and H. A.
Simon, Expert and novice performance in solving physics
problems, Science 208, 1335 (1980).
[84] M. T. H. Chi, P. Feltovich, and R. Glaser, Categorization
and representation of physics problems by experts and
novices, Cogn. Sci. 5, 121 (1981).
[85] P. T. Hardiman, R. Dufresne, and J. P. Mestre, The
relationship between problem categorization and problem
solving among experts and novices, Mem. Cogn. 17, 627
(1989).
[86] E. R. Savelsbergh, T. de Jong, and M. G. M. FergusonHessler, Choosing the right solution approach: The
crucial role of situational knowledge in electricity and
magnetism, Phys. Rev. ST Phys. Educ. Res. 7, 010103
(2011).
[87] M. T. H. Chi, M. Bassok, M. W. Lewis, P. Reimann, and
R. Glaser, Self-explanations: How students study and use
examples in learning to solve problems, Cogn. Sci. 13,
145 (1989).
[88] M. G. M. Ferguson-Hessler, and T. de Jong, Studying
physics texts: Differences in study processes between
good and poor performers, Cognit. Instr. 7, 41 (1990).
[89] E. Cohen, A. Mason, C. Singh, and E. Yerushalmi,
Identifying differences in diagnostic skills between physics students: Students self-diagnostic performance given
alternative scaffolding, AIP Conf. Proc. 1064, 99 (2008).
[90] E. Yerushalmi, A. Mason, E. Cohen, and C. Singh, Effects
of self-diagnosis on subsequent problem solving performance, AIP Conf. Proc. 1064, 53 (2008).
[91] M. Ward and J. Sweller, Structuring effective worked

examples, Cognit. Instr. 7, 1 (1990).
[92] J. I. Heller and F. Reif, Prescribing effective human
problem-solving processes: Problem description in physics, Cognit. Instr. 1, 177 (1984).
[93] J. H. Larkin, in Mental Models, edited by D. Gentner and
A. L. Stevens (Lawrence Erlbaum Assoc., Hillsdale, NJ,
1983), pp. 7598.
[94] B. Ibrahim and N. S. Rebello, Representational task
formats and problem solving strategies in kinematics
and work, Phys. Rev. ST Phys. Educ. Res. 8, 010126
(2012).
[95] D. Rosengrant, A. Van Heuvelen, and E. Etkina, Freebody diagrams: Necessary or sufficient? AIP Conf. Proc.
790, 177 (2005).
[96] A. Van Heuvelen and X. Zou, Multiple representations of
work-energy processes, Am. J. Phys. 69, 184 (2001).
[97] M. De Cock, Representation use and strategy choice in
physics problem solving, Phys. Rev. ST Phys. Educ. Res.
8, 020117 (2012).
[98] P. B. Kohl and N. D. Finkelstein, Student representational
competence and self-assessment when solving problems,
[99] D. E. Meltzer, Relation between students problemsolving performance and representational format, Am.
J. Phys. 73, 463 (2005).
[100] T. J. Bing and E. F. Redish, Epistemic complexity and the
journeyman-expert transition, Phys. Rev. ST Phys. Educ.
Res. 8, 010105 (2012).
[101] M. Bodin, Mapping university students epistemic framing of computational physics using network analysis,
[102] E. T. Torigoe and G. E. Gladding, Connecting symbolic
difficulties with failure in physics, Am. J. Phys. 79, 133
(2011).
[103] B. L. Sherin, The symbolic basis of physical intuition:
A study of two symbol systems in physics instruction,
Ph.D. thesis, University of California, Berkeley, 1996
(unpublished).
[104] B. L. Sherin, How students understand physics equations,
Cognit. Instr. 19, 479 (2001).
[105] E. Torigoe, What kind of math matters? A study of the
relationship between mathematical ability and success in
physics, Ph.D. thesis, University of Illinois at UrbanaChampaign, 2008 (unpublished).
[106] J. Tuminaro and E. F. Redish, Elements of a cognitive
model of physics problem solving: Epistemic games,
[107] N. L. Nguyen and D. E. Meltzer, Initial understanding of
vector concepts among students in introductory physics
courses, Am. J. Phys. 71, 630 (2003).
[108] L. Cui, N. S. Rebello, and A. G. Bennett, College
students transfer from calculus to physics, AIP Conf.
Proc. 818, 37 (2006).
[109] E. Cohen and S. E. Kanim, Factors influencing the
algebra reversal error, Am. J. Phys. 73, 1072 (2005).
[110] P. Heller and M. Hollabaugh, Teaching problem solving
through cooperative grouping. Part 2: Designing
problems and structuring groups, Am. J. Phys. 60, 637
(1992).
020119-46
[111] J. Mestre, Probing adults conceptual understanding and

transfer of learning via problem posing, J. Appl. Dev.
Psychol. 23, 9 (2002).
[112] T. L. OKuma, S. P. Maloney, and C. J. Hieggelke,
Ranking Task Exercises in Physics (Prentice-Hall, Upper
[113] A. Van Heuvelen, Experiment problems for mechanics,
Phys. Teach. 33, 176 (1995).
[114] A. Van Heuvelen and D. Maloney, Playing physics
jeopardy, Am. J. Phys. 67, 252 (1999).
[115] D. S. Wright and C. D. Williams, A WISE strategy for
introductory physics, Phys. Teach. 24, 211 (1986).
[116] C. Becerra-Labra, A. Gras-Mart, and J. Martnez
Torregrosa, The effects of a problem-based structure of
physics contents on conceptual learning and the ability to
solve problems, Int. J. Sci. Educ. 34, 1235 (2012).
[117] W. J. Leonard, R. J. Dufresne, and J. P. Mestre, Using
qualitative problem-solving strategies to highlight the role
of conceptual knowledge in solving problems, Am. J.
Phys. 64, 1495 (1996).
[118] A. Van Heuvelen, Overview, case study physics, Am. J.
Phys. 59, 898 (1991).
[119] P. Heller, R. Keith, and S. Anderson, Teaching problem
solving through cooperative grouping. Part 1: Group
versus individual problem solving, Am. J. Phys. 60,
627 (1992).
[120] G. Kortemeyer, E. Kashy, W. Benenson, and W. Bauer,
Experiences using the open-source learning content
management and assessment system LON-CAPA in
introductory physics courses, Am. J. Phys. 76, 438
(2008).
[121] Y.-J. Lee, D. J. Palazzo, R. Warnakulassoriya, and D. E.
Pritchard, Measuring student learning with item response
theory, Phys. Rev. ST Phys. Educ. Res. 4, 010102 (2008).
[122] A. Newell, J. D. Shaw, and H. A. Simon, Elements of a
theory of human problem solving, Psychol. Rev. 65, 151
(1958).
[123] A. Newell and H. A. Simon, Human Problem Solving
(Prentice-Hall, Englewood Cliffs, NJ, 1972).
[124] H. A. Simon, in Handbook of Learning and Cognitive
Processes, edited by W. K. Estes (Erlbaum, Hillsdale, NJ,
1978), Vol. 5, pp. 271295.
[125] M. L. Gick, Problem-solving strategies, Educ. Psychol.
21, 99 (1986).
[126] J. Sweller, Cognitive load during problem solving: Effects
on learning, Cogn. Sci. 12, 257 (1988).
[127] D. Gentner, Structure-mapping: A theoretical framework
for analogy, Cogn. Sci. 7, 155 (1983).
[128] D. Gentner, in Similarity and Analogical Reasoning, edited
by S. Vosniadou and A. Ortony (Cambridge University
Press, Cambridge, England, 1989), pp. 199241.
[129] K. Holyoak, in The Psychology of Learning and
Motivation, edited by G. H. Bower (Academic Press,
San Diego, CA, 1985), Vol. 19.
[130] K. J. Holyoak and K. Koh, Surface and structural
similarity in analogical transfer, Mem. Cogn. 15, 332
(1987).
[131] B. H. Ross, This is like that: The use of earlier problems
and the separation of similarity effects, J. Exp. Psychol.
Learn. Mem. Cogn. 13, 629 (1987).
[132] B. H. Ross, Distinguishing types of superficial similarities: Different effects on the access and use of earlier
problems, J. Exp. Psychol. Learn. Mem. Cogn. 15, 456
(1989).
[133] L. M. Reeves and R. W. Weisberg, The role of content and
abstract information in analogical transfer, Psychol. Bull.
115, 381 (1994).
[134] S. Y. Lin and C. Singh, Using isomorphic problems to
learn introductory physics, Phys. Rev. ST Phys. Educ.
Res. 7, 020104 (2011).
[135] D. Hammer, A. Elby, R. Scherr, and E. F. Redish, in
Transfer of Learning from a Modern Multidisciplinary
Perspective, edited by J. Mestre (Information Age
Publishing, Greenwich, CT, 2005), pp. 89119.
[136] J. Tuminaro, A cognitive framework for analyzing and
describing introductory students use and understanding of
mathematics in physics, Ph.D. thesis, University of
Maryland, 2004 (unpublished).
[137] J. S. Brown, A. Collins, and P. Duguid, Situated
cognition and the culture of learning, Educ. Res. 18,
32 (1989).
[138] A. Mason and C. Singh, Revisiting categorization, in
Proceedings of the 2009 National Association for
Research in Science Teaching (NARST) Annual Meeting
(Garden Grove, CA, 2009) Symposium S2.2, pp. 321.
[139] F. Mateycik, D. Jonassen, and N. S. Rebello, Using
Similarity Rating Tasks to Assess Case Reuse in Problem
Solving, Proceedings of the Physics Education Research
Conference (Ann Arbor, MI, 2010).[AIP Conference
Proceedings 1179, 201, (2009)].
[140] S. F. Wolf, D. P. Dougherty, and G. Kortemeyer, Empirical approach to interpreting card-sorting data, Phys. Rev.
ST Phys. Educ. Res. 8, 010124 (2012).
[141] A. Carmichael, A. Larson, E. Gire, L. Loschky, and N. S.
Rebello, How does visual attention differ between experts
and novices on physics problems?, AIP Conf. Proc. 1289,
93 (2010).
[142] D. Rosengrant, A. Van Heuvelen, and E. Etkina, Do
students use and understand free-body diagrams?, Phys.
[143] A. D. Smith, J. P. Mestre, and B. H. Ross, Eye gaze
patterns as students study worked out examples in
mechanics, Phys. Rev. ST Phys. Educ. Res. 6, 020118
(2010).
[144] T. de Jong and M. G. M. Ferguson-Hessler, Cognitive
structures of good and poor novice problem solvers in
physics, J. Educ. Psychol. 78, 279 (1986).
[145] J. L. Docktor, J. P. Mestre, and B. H. Ross, Impact of a
short intervention on novices categorization criteria,
[146] S. F. Wolf, D. P. Dougherty, and G. Kortemeyer, Rigging
the deck: Selecting good problems for expert-novice
card-sorting experiments, Phys. Rev. ST Phys. Educ.
Res. 8, 020116 (2012).
[147] M. Finegold and R. Mass, Differences in the process of
solving physics problems between good problem solvers
and poor problem solvers, Res. Sci. Technol. Educ. 3, 59
(1985).
[148] L. N. Walsh, R. G. Howard, and B. Bowe, Phenomenographic study of students problem solving approaches in
020119-47
[149]
[150]
[151]
[152]
[153]
[154]
[155]
[156]
[157]
[158]
[159]
[160]
[161]
[162]
[163]
[164]
[165]
[166]
physics, Phys. Rev. ST Phys. Educ. Res. 3, 020108

(2007).
A. Pawl, A Barrantes, and D. E. Pritchard, What do
seniors remember from freshman physics?, Phys. Rev. ST
Phys. Educ. Res. 8, 020118 (2012).
M. T. H. Chi, in The Cambridge Handbook of Expertise
and Expert Performance, edited by K. A. Ericsson, N.
Charness, R. R. Hoffman, and P. J. Feltovich (Cambridge
University Press, New York, 2006), pp. 2130.
F. Reif and J. I. Heller, Knowledge structure and problem
solving in physics, Educ. Psychol. 17, 102 (1982).
C. Singh, When physical intuition fails, Am. J. Phys. 70,
1103 (2002).
M. G. M. Ferguson-Hessler and T. de Jong, On the quality
of knowledge in the field of electricity and magnetism,
Am. J. Phys. 55, 492 (1987).
C. Henderson and K. A. Harper, Quiz corrections:
Improving learning by encouraging students to reflect
on their mistakes, Phys. Teach. 47, 581 (2009).
D. Rosengrant and T. Mzoughi, Preliminary study of
impulse-momentum diagrams, AIP Conf. Proc. 1064, 187
(2008).
C. J. De Leone and E. Gire, Is instructional emphasis on
the use of nonmathematical representations worth the
effort?, AIP Conf. Proc. 818, 45 (2006).
R. Ploetzner, E. Fehse, C. Kneser, and H. Spada, Learning
to relate qualitative and quantitative problem representations in a model-based setting for collaborative problem
solving, J. Learn. Sci. 8, 177 (1999).
D. Rosengrant, A. Van Heuvelen, and E. Etkina, Case
study: Students use of multiple representations in problem solving, AIP Conf. Proc. 818, 49 (2006).
P. B. Kohl and N. D. Finkelstein, Patterns of multiple
representation use by experts and novices during physics
problem solving, Phys. Rev. ST Phys. Educ. Res. 4,
010111 (2008).
J. M. Aguirre and G. Erickson, Students conceptions
about the vector characteristics of three physics concepts,
J. Res. Sci. Teach. 21, 439 (1984).
S. Flores, S. E. Kanim, and C. H. Kautz, Student use of
vectors in introductory mechanics, Am. J. Phys. 72, 460
(2004).
P. S. Shaffer and L. C. McDermott, A research-based
approach to improving student understanding of the
vector nature of kinematical concepts, Am. J. Phys. 73,
921 (2005).
K. Christianson, J. P. Mestre, and S. G. Luke, Practice
makes (nearly) perfect: Solving studentsand
professorstype algebra word problems, Appl. Cogn.
Psychol. 26, 810 (2012).
M. Bassok and K. J. Holyoak, Interdomain transfer
between isomorphic topics in algebra and physics,
J. Exp. Psychol. Learn. Mem. Cogn. 15, 153
(1989).
D. C. Meredith and K. A. Marrongelle, How students use
mathematical resources in an electrostatics context, Am.
J. Phys. 76, 570 (2008).
M. S. Sabella and E. F. Redish, Knowledge organization
and activation in physics problem solving, Am. J. Phys.
75, 1017 (2007).
[167] T. J. Bing and E. F. Redish, Analyzing problem solving

using math in physics: Epistemological framing via
warrants, Phys. Rev. ST Phys. Educ. Res. 5, 020108
(2009).
[168] E. C. Sayre and M. C. Wittmann, Plasticity of intermediate mechanics students coordinate system choice, Phys.
[169] C. A. Ogilvie, Changes in students problem-solving
strategies in a course that includes context-rich, multifaceted problems, Phys. Rev. ST Phys. Educ. Res. 5,
020102 (2009).
[170] A. Van Heuvelen, L. D. Allen, and P. Mihas, Experiment
problems for electricity and magnetism, Phys. Teach. 37,
482 (1999).
[171] D. P. Maloney, Ranking tasks: A new type of test item,
J. Coll. Sci. Teach. 16, 510 (1987).
[172] See http://relate.mit.edu/teaching/realworldproblems/.
[173] See http://www.physics.umd.edu/perg/abp/think/.
[174] L. Ding, N. Reay, A. Lee, and L. Bao, Exploring the
role of conceptual scaffolding in solving synthesis
problems, Phys. Rev. ST Phys. Educ. Res. 7, 020109
(2011).
[175] D. Huffman, Effect of explicit problem solving instruction on high school students problem-solving performance and conceptual understanding of physics, J. Res.
Sci. Teach. 34, 551 (1997).
[176] J. H. P. Van Weeren, F. F. M. de Mul, M. J. Peters, H.
Kramers-Pals, and H. J. Roossink, Teaching problemsolving in physics: A course in electromagnetism, Am. J.
Phys. 50, 725 (1982).
[177] G. Plya, How to Solve It (Princeton University Press,
Princeton, NJ, 1945).
[178] K. Heller and P. J. Heller, Competent Problem Solver
Calculus Version (McGraw-Hill, New York, 2000).
[179] F. Reif, Understanding Basic Mechanics, Text and Workbook (Wiley, New York, 1995).
[180] R. Gautreau and L. Novemsky, Concepts firstA small
group approach to physics learning, Am. J. Phys. 65, 418
(1997).
[181] R. J. Dufresne, W. J. Gerace, P. T. Hardiman, and J. P.
Mestre, Constraining novices to perform expertlike problem analyses: Effects on schema acquisition, J. Learn. Sci.
2, 307 (1992).
[182] J. P. Mestre, R. J. Dufresne, W. J. Gerace, P. T. Hardiman,
and J. S. Touger, Promoting skilled problem-solving
behavior among beginning physics students, J. Res.
Sci. Teach. 30, 303 (1993).
[183] A. Van Heuvelen, ALPS: Mechanics (Hayden-McNeil
Publishing, Plymouth, MI 1994), Vol. 1.
[184] A. Van Heuvuelen, ALPS: Electricity and Magnetism
(Hayden-McNeil Publishing, Plymouth, MI, 1994),
Vol. 2.
[185] R. Taconis, M. G. M. Ferguson-Hessler, and H.
Broekkamp, Teaching science problem-solving: An overview of experimental work, J. Res. Sci. Teach. 38, 442
(2001).
[186] D. Kane and B. Sherwood, A computer-based course in
classical mechanics, Comput. Educ. 4, 15 (1980).
[187] B. A. Sherwood, Free-body diagrams (a PLATO lesson),
Am. J. Phys. 39, 1199 (1971).
020119-48
[188] S. G. Smith and B. A. Sherwood, Educational uses of the

PLATO computer system, Science 192, 344 (1976).
[189] A. S. Gertner and K. VanLehn, in Proceedings of the 5th
International Conference, ITS 2000, edited by G.
Gautheier, C. Frasson, and K. VanLehn (Springer,
New York, 2000), pp. 133142.
[190] Q. Ryan, E. Frodermann, K. Heller, L. Hsu, and B. Aryal,
Designing a useful problem solving coach: Usage and
usability studies, in Proceedings of the Physics Education
Research Conference 2013, Portland, OR, 2014 edited
by P. V. Engelhardt, A. D. Churukian, D. L. Jones
pp. 309312.
[191] E. Kashy, B. M. Sherrill, Y. Tsai, D. Thaler, D. Weinshank,
M. Engelmann, and D. J. Morrissey, CAPA, an integrated
computer assisted personalized assignment system, Am. J.
Phys. 61, 1124 (1993).
[192] E. Kashy, S. J. Gaff, N. H. Pawley, W. L. Stretch, S. L.
Wolfe, D. J. Morrissey, and Y. Tsai, Conceptual questions
in computer-assisted assignments, Am. J. Phys. 63, 1000
(1995).
[193] A. M. Pascarella, CAPA (Computer-Assisted Personalized Assignments) in a Large University Setting,
Ph.D. thesis, University of Colorado, Boulder, 2002
(unpublished).
[194] S. W. Bonham, D. L. Deardorff, and R. J. Beichner,
Comparison of student performance using web and
paper-based homework in college level physics, J. Res.
Sci. Teach. 40, 1050 (2003).
[195] K. K. Cheng, B. A. Thacker, R. L. Cardenas, and C.
Crouch, Using an online homework system enhances
students learning of physics concepts in an introductory
physics course, Am. J. Phys. 72, 1447 (2004).
[196] J. Mestre, D. M. Hart, K. A. Rath, and R. Dufresne, The
effect of web-based homework on test performance in
large enrollment introductory physics courses, J. Comput.
Math. Sci. Teach. 21, 229 (2002).
[197] E.-S. Morote and D. E. Pritchard, What course elements
correlate with improvement on tests in introductory
Newtonian mechanics?, Am. J. Phys. 77, 746 (2009).
[198] K. VanLehn, C. Lynch, K. Schulze, J. A. Shapiro, R.
Shelby, L. Taylor, D. Treacy, A. Weinstein, and
M. Wintersgill, The Andes physics tutoring system:
Lessons learned, Int. J. Artif. Intell. Educ. 15, 147
(2005).
[199] E. Gire, Between the poles: Locating physics majors in
the expert-novice continuum Ph.D. thesis, University of
California, San Diego, 2007 (unpublished).
[200] K. A. Harper, Investigating the development of problem
solving skills during a freshman physics sequence Ph.D.
thesis, The Ohio State University, 2001 (unpublished).
[201] C. Henderson, E. Yerushalmi, V. Kuo, P. Heller, and K.
Heller, Grading student problem solutions: The challenge
of sending a consistent message, Am. J. Phys. 72, 164
(2004).
[202] W. K. Adams and C. E. Wieman, Problem solving skill
evaluation instrumentValidation studies, AIP Conf.
Proc. 883, 18 (2006).
[203] J. D. Marx and K. Cummings, in Development of a survey
instrument to gauge students problem-solving abilities,
AIP Conf. Proc. 1289 221 (2010).
[204] J. L. Docktor, Development and validation of a

physics problem-solving assessment rubric, Ph.D. thesis,
University of Minnesota, 2009 (unpublished).
[205] E. F. Redish, Teaching Physics with the Physics Suite
(John Wiley & Sons, Hoboken, NJ, 2003).
[206] A. B. Arons, A Guide to Introductory Physics Teaching
(John Wiley & Sons, New York, 1990).
[207] R. D. Knight, Five Easy Lessons: Strategies for Successful Physics Teaching (Addison-Wesley, Glenview, IL,
2002).
[208] D. E. Meltzer and R. K. Thornton, Resource letter
ALIP-1: Active-learning instruction in physics, Am. J.
Phys. 80, 478 (2012).
[209] R. R. Hake, Interactive-engagement versus traditional
methods: A six-thousand student survey of mechanics
test data for introductory physics courses, Am. J. Phys.
66, 64 (1998).
[210] D. E. Meltzer and K. Manivannan, Transforming the
lecture-hall environment: The fully interactive physics
lecture, Am. J. Phys. 70, 639 (2002).
[211] L. Ding, N. W. Reay, N. Lee, and L. Bao, Are we
asking the right questions? Validating clicker question
sequences by student interviews, Am. J. Phys. 77, 643
(2009).
[212] R. J. Dufresne, W. J. Gerace, W. J. Leonard, J. P. Mestre,
and L. Wenk, Classtalk: A classroom communication
system for active learning, J. Comput. High Educ. 7, 3
(1996).
[213] E. Mazur, Peer Instruction: A Users Manual (PrenticeHall, Upper Saddle River, NJ, 1997).
[214] D. R. Sokoloff and R. K. Thornton, Interactive Lecture
Demonstrations, Active Learning in Introductory Physics
(John Wiley & Sons, Hoboken, NJ, 2004).
[215] G. M. Novak, E. T. Patterson, A. D. Gavrin, and W.
Christian, Just in Time Teaching (Prentice-Hall, Upper
[216] Z. Chen, T. Stelzer, and G. Gladding, Using multimedia
modules to better prepare students for introductory
physics lecture, Phys. Rev. ST Phys. Educ. Res. 6,
010108 (2010).
[217] T. Stelzer, D. T. Brookes, G. Gladding, and J. P. Mestre,
Impact of multimedia learning modules on an introductory course on electricity and magnetism, Am. J. Phys. 78,
755 (2010).
[218] T. Stelzer, G. Gladding, J. P. Mestre, and D. T. Brookes,
Comparing the efficacy of multimedia modules with
traditional textbooks for learning introductory physics
content, Am. J. Phys. 77, 184 (2009).
[219] E. Etkina, Weekly reports: A two-way feedback tool, Sci.
Educ. 84, 594 (2000).
[220] K. A. Harper, E. Etkina, and Y. Lin, Encouraging and
analyzing student questions in a large physics course:
Meaningful patterns for instructors, J. Res. Sci. Teach. 40,
776 (2003).
[221] D. B. May and E. Etkina, College physics students
epistemological self-reflection and its relationship to
conceptual learning, Am. J. Phys. 70, 1249 (2002).
[222] H. G. Close and P. R. L. Heron, Student understanding of
the angular momentum of classical particles, Am. J. Phys.
79, 1068 (2011).
020119-49
[223] S. J. Pollock and N. D. Finkelstein, Sustaining educational reforms in introductory physics, Phys. Rev. ST
Phys. Educ. Res. 4, 010110 (2008).
[224] A. Elby, Helping physics students learn how to learn, Am.
J. Phys. 69, S54 (2001).
[225] M. C. Wittmann, R. N. Steinberg, E. F. Redish, and the
University of Maryland Physics Education Research
Group, Activity-Based Tutorials Volume 1: Introductory
Physics, The Physics Suite (John Wiley & Sons,
New York, 2004).
[226] M. C. Wittmann, R. N. Steinberg, E. F. Redish, and the
University of Maryland Physics Education Research
Group, Activity-Based Tutorials Volume 2: Modern
Physics (John Wiley & Sons, New York, 2004).
[227] D. R. Sokoloff, R. K. Thornton, and P. W. Laws,
RealTime Physics Active Learning Laboratories Module
1: Mechanics, 2nd ed. (John Wiley & Sons, Hoboken, NJ,
2004).
2: Heat and Thermodynamics, 2nd ed. (John Wiley &
Sons, Hoboken, NJ, 2004).
3: Electric Circuits, 2nd ed. (John Wiley & Sons,
Hoboken, NJ, 2004).
4: Light and Optics, 2nd ed. (John Wiley & Sons,
Hoboken, NJ, 2004).
[231] R. J. Beichner, The impact of video motion analysis on
kinematics graph interpretation skills, Am. J. Phys. 64,
1272 (1996).
[232] E. Etkina, and A. Van Heuvelen, Investigative Science
Learning Environment A science process approach to
learning physics, in Reviews in PER Volume 1: ResearchBased Reform of University Physics, edited by E. F.
Redish and P. J. Cooney (American Association of
Physics Teachers, College Park, MD, 2007).
[233] E. Etkina, A. Karelina, M. Ruibal-Villasenor, D. Rosengrant,
R. Jordan, and C. E. Hmelo-Silver, Design and reflection
help students develop scientific abilities: Learning in
introductory physics laboratories, J. Learn. Sci. 19, 54
(2010).
[234] L. McCullough, The effect of introducing computers
into an introductory problem-solving laboratory,
Ph.D. thesis, University of Minnesota, 2000
(unpublished).
[235] R. Lippmann Kung, Teaching the concepts of measurement: An example of a concept-based laboratory course,
Am. J. Phys. 73, 771 (2005).
[236] B. J. Duch, S. E. Groh, and D. E. Allen, The Power of
Problem-Based Learning (Stylus Publishing, Sterling,
VA, 2001).
[237] M. A. Pease and D. Kuhn, Experimental analysis of the
effective components of problem-based learning, Sci.
Educ. 95, 57 (2011).
[238] L. C. McDermott and The Physics Education Group at the
University of Washington, Physics by Inquiry (John
Wiley & Sons, New York, 1996), Vol. I.
[239] L. C. McDermott and The Physics Education Group at the

University of Washington, Physics by Inquiry (John
Wiley & Sons, New York, 1996), Vol. II.
[240] P. Laws, Calculus-based physics without lectures, Phys.
Today 44, 24 (1991).
[241] P. W. Laws, Workshop Physics Activity Guide (John
Wiley & Sons, New York, 2004), Vols. 14.
[242] K. Cummings, J. Marx, R. Thornton, and D. Kuhl,
Evaluating innovation in studio physics, Am. J. Phys.
67, S38 (1999).
[243] C. Hoellwarth, M. Moelter, and R. Knight, A direct
comparison of conceptual learning and problem solving
ability in traditional and studio style classrooms, Am. J.
Phys. 73, 459 (2005).
[244] P. B. Kohl and H. V. Kuo, Chronicling a successful
secondary implementation of Studio Physics, Am. J.
Phys. 80, 832 (2012).
[245] C. M. Sorensen, A. D. Churukian, S. Maleki, and
D. A. Zollman, The New Studio format for instruction
of introductory physics, Am. J. Phys. 74, 1077 (2006).
[246] J. Wilson, The CUPLE physics studio, Phys. Teach. 32,
518 (1994).
[247] R. J. Beichner, J. M. Saul, D. S. Abbott, J. J. Morse, D. L.
Deardorff, R. J. Allain, S. W. Bonham, M. H. Dancy, and
J. S. Risley, The student-centered activities for large
enrollment undergraduate programs (SCALE-UP)
project, in, Reviews in PER Volume 1: Research-Based
Reform of University Physics, edited by E. F. Redish and
P. J. Cooney (American Association of Physics Teachers,
College Park, MD, 2007).
[248] J. Belcher, Studio physics at MIT, MIT Physics
Annual 2001, 58 (2001), http://web.mit.edu/physics/
papers/Belcher_physicsannual_fall_01.pdf,
retrieved
October 6, 2010.
[249] Y. J. Dori and J. Belcher, How does technology-enabled
active learning affect undergraduate students understanding of electromagnetism concepts?, J. Learn. Sci. 14, 243
(2005).
[250] Y. J. Dori, J. Belcher, M. Bessette, M. Danziger, A.
McKinney, and E. Hult, Technology for active learning,
Mater. Today 6, 44 (2003).
[251] K. Cummings, P. Laws, E. Redish, and P. Cooney,
Understanding Physics (John Wiley & Sons, Hoboken,
NJ, 2004).
[252] R. W. Chabay and B. A. Sherwood, Matter and Interactions, 3rd ed. (John Wiley & Sons, Hoboken, NJ,
2009).
[253] W. K. Adams, S. Reid, R. LeMaster, S. B. McKagan,
K. K. Perkins, M. Dubson, and C. E. Wieman, A study of
educational simulations Part I: Engagement and learning,
J. Interact. Learn. Res. 19, 397 (2008).
[254] W. K. Adams, S. Reid, R. LeMaster, S. B. McKagan,
K. K. Perkins, M. Dubson, and C. E. Wieman, A study of
educational simulations Part II: Interface design, J.
Interact. Learn. Res. 19, 551 (2008).
[255] N. D. Finkelstein, W. K. Adams, C. J. Keller, P. B. Kohl,
K. K. Perkins, N. S. Podolefsky, S. Reid, and R.
LeMaster, When learning about the real world is better
done virtually: A study of substituting computer
020119-50
[256]
[257]
[258]
[259]
[260]
[261]
[262]
[263]
[264]
[265]
[266]
[267]
[268]
[269]
[270]
[271]
[272]
[273]
simulations for laboratory equipment, Phys. Rev. ST

Phys. Educ. Res. 1, 010103 (2005).
R. Chabay and B. Sherwood, Computational physics in
the introductory calculus-based course, Am. J. Phys. 76,
307 (2008).
E. F. Redish and J. M. Wilson, Student programming in
the introductory physics course: M.U.P.P.E.T., Am. J.
Phys. 61, 222 (1993).
I. Halloun and D. Hestenes, The initial knowledge
state of college physics students, Am. J. Phys. 53,
1043 (1985).
W. J. Leonard, R. J. Dufresne, W. J. Gerace, and J. P.
Mestre, Minds-On Physics: Activities and Reader
(Kendall-Hunt Publishing, Dubuque, IA, 2000), Vols. 16.
E. Brewe, Modeling theory applied: Modeling Instruction in introductory physics, Am. J. Phys. 76, 1155
(2008).
D. H. Jonassen and S. M. Land, Theoretical Foundations
of Learning Environments (Lawrence Erlbaum Assoc.,
Mahweh, NJ, 2000).
L. W. Trowbridge, R. W. Bybee, and J. C. Powell, Teaching Secondary School Science: Strategies for Developing
Scientific Literacy, 7th ed. (Prentice-Hall, Upper Saddle
River, NJ, 2000).
E. Von Glasersfeld, Cognition, construction of knowledge, and teaching,, Synthese 80, 121 (1989).
J. Dewey, Experience and Education, 60th anniversary
ed. (Kappa Delta Pi, Indianapolis, IN, 1998).
D. Jonassen, Designing constructivist learning environments, in Instructional-Design Theories and Models: A
New Paradigm of Instructional Theory, edited by C. M.
Reigeluth (Lawrence Erlbaum Assoc., Mahwah, NJ,
1999), Vol. II, pp. 217240.
R. Karplus, Science teaching and the development of
reasoning, J. Res. Sci. Teach. 14, 169 (1977).
D. Zollman, in Proceedings of ICUPE, College Park,
MD, 1996, edited by E. F. Redish and J. S. Rigden
(American Institute of Physics, Melville, NY, 1997),
pp. 11371149.
C. A. Manogue, P. J. Siemens, J. Tate, K. Browne, M. L.
Niess, and A. J. Wolfer, Paradigms in Physics: A
new upper-division curriculum, Am. J. Phys. 69, 978
(2001).
D. E. Trowbridge and L. C. McDermott, Investigation of
student understanding of the concept of velocity in one
dimension, Am. J. Phys. 48, 1020 (1980).
P. S. Shaffer and L. C. McDermott, Research as a guide
for curriculum development: An example from introductory electricity. Part II: Design of instructional strategies,
Am. J. Phys. 60, 1003 (1992).
D. E. Trowbridge and L. C. McDermott, Investigation of
student understanding of the concept of acceleration in
one dimension, Am. J. Phys. 49, 242 (1981).
E. Etkina, A. Van Heuvelen, S. White-Brahmia, D. T.
Brookes, M. Gentile, S. Murthy, D. Rosengrant, and A.
Warren, Scientific abilities and their assessment, Phys.
C. Crouch and E. Mazur, Peer Instruction: Ten years
of experience and results, Am. J. Phys. 69, 970
(2001).
[274] N. Lasry, E. Mazur, and J. Watkins, Peer Instruction:

From Harvard to the two-year college, Am. J. Phys. 76,
1066 (2008).
[275] I. D. Beatty, W. J. Gerace, W. J. Leonard, and R. J.
Dufresne, Designing effective questions for classroom
response system teaching, Am. J. Phys. 74, 31 (2006).
[276] C. Keller, N. D. Finkelstein, K. Perkins, S. J. Pollock, C.
Turpen, and M. Dubson, Research-based practices for
effective clicker use, AIP Conf. Proc. 951, 128 (2007).
[277] M. C. James, The effect of grading incentive on student
discourse in Peer Instruction, Am. J. Phys. 74, 689
(2006).
[278] For more information, see http://www.iclicker.com/dnn/;
http://www.einstruction.com/.
[279] J. E. Caldwell, Clickers in the large classroom: Current
research and best-practice tips, CBE Life Sci. Educ. 6, 9
(2007).
[280] N. Lasry, Clickers or flashcards: Is there really a difference?, Phys. Teach. 46, 242 (2008).
[281] C. H. Crouch, A. P. Fagen, J. P. Callan, and E. Mazur,
Classroom demonstrations: Learning tools or entertainment?, Am. J. Phys. 72, 835 (2004).
[282] S. P. Formica, J. L. Easley, and M. C. Spraker,
Transforming common-sense beliefs into Newtonian
thinking through Just-In-Time Teaching, Phys. Rev. ST
Phys. Educ. Res. 6, 020106 (2010).
[283] See
http://research.physics.illinois.edu/per/prelectures
.html.
[284] H. R. Sadaghiani, Using multimedia learning modules in
a hybrid-online course in electricity and magnetism, Phys.
[285] H. R. Sadaghiani, Controlled study on the effectiveness of
multimedia learning modules for teaching mechanics,
[286] M. L. Rosenquist and L. C. McDermott, A conceptual
approach to teaching kinematics, Am. J. Phys. 55, 407
(1987).
[287] L. C. McDermott, M. L. Rosenquist, and E. H. van Zee,
Student difficulties in connecting graphs and physics:
Examples from kinematics, Am. J. Phys. 55, 503
(1987).
[288] L. C. McDermott, P. S. Shaffer, and M. D. Somers,
Research as a guide for teaching introductory mechanics:
An illustration in the context of the Atwoods machine,
Am. J. Phys. 62, 46 (1994).
[289] R. A. Lawson and L. C. McDermott, Student understanding of the work-energy and impulse-momentum theorems, Am. J. Phys. 55, 811 (1987).
[290] T. OBrien Pride, S. Vokos, and L. C. McDermott, The
challenge of matching learning assessments to teaching
goals: An example from the work-energy and impulsemomentum theorems, Am. J. Phys. 66, 147 (1998).
[291] F. M. Goldberg and L. C. McDermott, An investigation of
student understanding of the real image formed by a
converging lens or concave mirror, Am. J. Phys. 55, 108
(1987).
[292] B. S. Ambrose, P. S. Shaffer, R. N. Steinberg, and L. C.
McDermott, An investigation of student understanding of
single-slit diffraction and double-slit interference, Am. J.
Phys. 67, 146 (1999).
020119-51
[293] S. Vokos, P. S. Shaffer, B. S. Ambrose, and L. C.

McDermott, Student understanding of the wave nature
of matter: Diffraction and interference of particles, Am. J.
Phys. 68, S42 (2000).
[294] K. Wosilait, P. R. L. Heron, P. S. Shaffer, and L. C.
McDermott, Addressing student difficulties in applying
a wave model to the interference and diffraction of light,
Am. J. Phys. 67, S5 (1999).
[295] C. H. Kautz, P. R. L. Heron, M. E. Loverude, and L. C.
McDermott, Student understanding of the ideal gas law,
Part I: A macroscopic perspective, Am. J. Phys. 73, 1055
(2005).
[296] C. H. Kautz, P. R. L. Heron, P. S. Shaffer, and L. C.
McDermott, Student understanding of the ideal gas
law, Part II: A microscopic perspective, Am. J. Phys.
73, 1064 (2005).
[297] M. E. Loverude, C. H. Kautz, and P. R. L. Heron, Student
understanding of the first law of thermodynamics: Relating work to the adiabatic compression of an ideal gas,
Am. J. Phys. 70, 137 (2002).
[298] M. J. Cochran and P. R. L. Heron, Development and
assessment of research-based tutorials on heat engines
and the second law of thermodynamics, Am. J. Phys. 74,
734 (2006).
[299] M. E. Loverude, P. R. L. Heron, and C. H. Kautz, Identifying and addressing student difficulties with hydrostatic
pressure, Am. J. Phys. 78, 75 (2010).
[300] L. G. Ortiz, P. R. L. Heron, and P. S. Shaffer, Student
understanding of static equilibrium: Predicting and
accounting for balancing, Am. J. Phys. 73, 545
(2005).
[301] H. G. Close and P. R. L. Heron, Research as a guide for
improving student learning: An example from momentum
conservation, Am. J. Phys. 78, 961 (2010).
[302] R. N. Steinberg, M. C. Wittmann, and E. F. Redish,
Mathematical tutorials in introductory physics, AIP Conf.
Proc. 399, 1075 (1997).
[303] N. D. Finkelstein and S. J. Pollock, Replicating and
understanding successful innovations: Implementing
tutorials in introductory physics, Phys. Rev. ST Phys.
Educ. Res. 1, 010101 (2005).
[304] H. V. Mauk and D. Hingley, Student understanding of
induced current: Using tutorials in introductory physics to
teach electricity and magnetism, Am. J. Phys. 73, 1164
(2005).
[305] E. F. Redish, J. M. Saul, and R. N. Steinberg, On the
effectiveness of active-engagement microcomputer-based
laboratories, Am. J. Phys. 65, 45 (1997).
[306] T. I. Smith and M. C. Wittmann, Comparing three
methods for teaching Newtons third law, Phys. Rev.
ST Phys. Educ. Res. 3, 020105 (2007).
[307] D. W. Johnson, R. T. Johnson, and E. J. Holubec,
Cooperation in the Classroom (Interaction Book
Company, Edina, MN, 1988).
[308] D. W. Johnson, R. T. Johnson, and K. A. Smith, Active
Learning: Cooperation in the College Classroom
(Interaction Book Company, Edina, MN, 1991).
[309] H. Brasell, The effect of real-time laboratory graphing on
learning graphic representations of distance and velocity,
J. Res. Sci. Teach. 24, 385 (1987).
[310] R. K. Thornton and D. R. Sokoloff, Learning motion

concepts using real-time microcomputer-based laboratory
tools, Am. J. Phys. 58, 858 (1990).
[311] R. Beichner, The effect of simultaneous motion presentation and graph generation in a kinematics lab, J. Res.
Sci. Teach. 27, 803 (1990).
[312] J. Brungardt and D. Zollman, Influence of interactive
videodisc instruction using simultaneous-time analysis on
kinematics graphing skills of high school physics students, J. Res. Sci. Teach. 32, 855 (1995).
[313] See
http://groups.physics.umn.edu/physed/Research/
CPSL/MBL.html.
[314] R. K. Thornton and D. R. Sokoloff, RealTime Physics:
Active learning laboratory, AIP Conf. Proc. 399, 1101
(1997).
[315] D. Zollman and R. Fuller, Teaching and learning physics
with interactive video, Phys. Today 47, 41 (1994).
[316] L. T. Escalada and D. A. Zollman, An investigation of the
effects of using interactive digital video in a physics
classroom on student learning and attitudes, J. Res. Sci.
Teach. 34, 467 (1997).
[317] E. Etkina, S. Murthy, and X. Zou, Using introductory labs
to engage students in experimental design, Am. J. Phys.
74, 979 (2006).
[318] W. M. Reid and D. F. Arseneau, Labs of unlimited scope,
Am. J. Phys. 39, 271 (1971).
[319] A. Van Heuvelen and E. Etkina, The Physics Active
Learning Guide (Addison-Wesley, San Francisco, 2006).
[320] See also http://www.islephysics.net/.
[321] A. Karelina and E. Etkina, Acting like a physicist: Student
approach study to experimental design, Phys. Rev. ST
Phys. Educ. Res. 3, 020106 (2007).
[322] See
http://groups.physics.umn.edu/physed/Research/
PSL/pslintro.html.
[323] See http://umdperg.pbworks.com/ScientificCommunity
Labs.
[324] For more information, see http://www.udel.edu/inst.
[325] B. J. Duch, Problem-based learning in physics: The power
of students teaching students, J. Coll. Sci. Teach. 15, 326
(1996).
[326] B. J. Duch, Problem-based learning in physics: Making
connections with the real world, AIP Conf. Proc. 399, 557
(1997).
[327] See also http://www.phys.washington.edu/groups/peg/pbi
.html.
[328] P. R. L. Heron, M. E. Loverude, P. S. Shaffer, and L. C.
McDermott, Helping students develop an understanding
of Archimedes principle. II. Development of researchbased instructional materials, Am. J. Phys. 71, 1188
(2003).
[329] B. Thacker, E. Kim, K. Trefz, and S. M. Lea, Comparing
problem solving performance of physics students in
inquiry-based and traditional introductory physics
courses, Am. J. Phys. 62, 627 (1994).
[330] See http://physics.dickinson.edu/~wp_web/WP_homepage
.html.
[331] See http://physics.dickinson.edu/~eip_web/EiP_homepage
.html.
[332] See http://www.ncsu.edu/per/scaleup.html.
[333] See http://scaleup.ncsu.edu/FAQs.html.
020119-52
[334] J. Gaffney, E. Richards, M. B. Kustusch, L. Ding, and R.

Beichner, Scaling up education reform, J. Coll. Sci.
Teach. 37, 48 (2008).
[335] Y. J. Dori and J. Belcher, in Visualization in Science
Education, edited by J. K. Gilbert (Springer, Amsterdam,
2005), pp. 187216.
[336] R. D. Knight, Physics for Scientists and Engineers: A
Strategic Approach with Modern Physics and Mastering
Physics (Addison-Wesley, San Francisco, 2004).
[337] T. A. Moore, Six Ideas That Shaped Physics (McGrawHill, New York, 2003).
[338] See http://www.physics.pomona.edu/sixideas/sisuc.html.
[339] R. W. Chabay and B. A. Sherwood, Bringing atoms into
first-year physics, Am. J. Phys. 67, 1045 (1999).
[340] R. Chabay and B. Sherwood, Modern mechanics, Am. J.
Phys. 72, 439 (2004).
[341] R. Chabay and B. Sherwood, Restructuring the introductory electricity and magnetism course, Am. J. Phys. 74,
329 (2006).
[342] M. A. Kohlmyer, M. D. Caballero, R. Catrambone, R. W.
Chabay, L. Ding, M. P. Haugan, M. J. Marr, B. A.
Sherwood, and M. F. Schatz, Tale of two curricula:
The performance of 2000 students in introductory electromagnetism, Phys. Rev. ST Phys. Educ. Res. 5, 020105
(2009).
[343] M. D. Caballero, E. F. Greco, E. R. Murray, K. R. Bujak,
and M. J. Marr, Comparing large lecture mechanics
curricula using the Force Concept Inventory: A five
thousand student study, Am. J. Phys. 80, 638 (2012).
[344] M. H. Dancy and R. Beichner, Impact of animation
on assessment of conceptual understanding in
physics, Phys. Rev. ST Phys. Educ. Res. 2, 010104
(2006).
[345] S. Yeo, R. Loss, M. Zadnik, A. Harrison, and D.
Treagust, What do students really learn from interactive
multimedia? A physics case study, Am. J. Phys. 72, 1351
(2004).
[346] R. N. Steinberg, Computers in teaching science: To
simulate or not to simulate?, Am. J. Phys. 68, S37
(2000).
[347] J. M. Casperson and M. C. Linn, Using visualizations to
teach electrostatics, Am. J. Phys. 74, 316 (2006).
[348] See, for example, ActivPhysics OnLine, http://www
.awbc.com/knight.
[349] K. Perkins, W. Adams, M. Dubson, N. Finkelstein, S.
Reid, C. Wieman, and R. LeMaster, PhET: Interactive
simulations for teaching and learning physics, Phys.
Teach. 44, 18 (2006).
[350] S. B. McKagan, K. K. Perkins, M. Dubson, C. Malley, S.
Reid, R. LeMaster, and C. E. Wieman, Developing and
researching PhET simulations for teaching quantum
mechanics, Am. J. Phys. 76, 406 (2008).
[351] Special double issue of the American Journal of Physics
on incorporating computation into the physics curriculum
and using computation to enhance physics teaching [Am.
J. Phys. 76, Issue No. 4 (2008)].
[352] L. Bao, S. R. Stonebraker, and H. Sadaghiani, A flexible
homework method, Am. J. Phys. 76, 878 (2008).
[353] See an overview of the research project and sample
materials at http://cpucips.sdsu.edu/petpset.html.
[354] V. K. Otero and K. E. Gray, Attitudinal gains across

multiple universities using the Physics and Everyday
Thinking curriculum, Phys. Rev. ST Phys. Educ. Res.
4, 020104 (2008).
[355] See
http://www.iat.com/courses/highschoolscience/
activephysics/.
[356] See http://cpucips.sdsu.edu/interactionsinfo/.
[357] D. H. McIntyre, J. Tate, and C. A. Manogue, Integrating
computational activities into the upper-level Paradigms in
Physics curriculum at Oregon State University, Am. J.
Phys. 76, 340 (2008).
[358] B. S. Ambrose, Investigating student understanding
in intermediate mechanics: Identifying the need for a
tutorial approach to instruction, Am. J. Phys. 72, 453
(2004).
[359] S. V. Chasteen and S. J. Pollock, A researchedbased approach to assessing student learning issues in
upper-division electricity and magnetism, AIP Conf.
Proc. 1179, 7 (2009).
[360] S. V. Chasteen, S. J. Pollock, R. E. Pepper, and K. K.
Perkins, Thinking like a physicist: A multi-semester case
study of junior-level electricity and magnetism, Am. J.
Phys. 80, 923 (2012).
[361] S. V. Chasteen, S. J. Pollock, R. E. Pepper, and K. K.
Perkins, Transforming the junior level: Outcomes from
instruction and research in E&M, Phys. Rev. ST Phys.
Educ. Res. 8, 020107 (2012).
[362] S. J. Pollock, S. V. Chasteen, M. Dubson, and K. K.
Perkins, The use of concept tests and peer instruction
in upper-division physics, AIP Conf. Proc. 1289, 261
(2010).
[363] S. Goldhaber, S. Pollock, M. Dubson, P. Beale, and K.
Perkins, Transforming upper-division quantum mechanics: Learning goals and assessment, AIP Conf. Proc.
1179, 145 (2009).
[364] L. D. Carr and S. B. McKagan, Graduate quantum
mechanics reform, Am. J. Phys. 77, 308 (2009).
[365] V. Otero, S. Pollock, and N. Finkelstein, A physic
departments role in preparing physics teachers: The
Colorado learning assistant model, Am. J. Phys. 78,
1218 (2010).
[366] R. M. Goertzen, E. Brewe, L. H. Kramer, L. Wells, and D.
Jones, Moving toward change: Institutionalizing reform
through implementation of the Learning Assistant model
and Open Source Tutorial, Phys. Rev. ST Phys. Educ.
Res. 7, 020105 (2011).
[367] See
http://www.aapt.org/Conferences/newfaculty/nfw
.cfm.
[368] C. Henderson and M. H. Dancy, Impact of physics
education research on the teaching of introductory quantitative physics in the United States, Phys. Rev. ST Phys.
Educ. Res. 5, 020107 (2009).
[369] M. Dancy and C. Henderson, Pedagogical practices and
instructional change of physics faculty, Am. J. Phys. 78,
1056 (2010).
[370] C. Singh, Student understanding of quantum mechanics,
Am. J. Phys. 69, 885 (2001).
[371] E. F. Redish and D. Hammer, Reinventing college physics
for biologists: Explicating an epistemological curriculum,
Am. J. Phys. 77, 629 (2009).
020119-53
[372] M. Plomer, K. Jessen, G. Rangelov, and M. Meyer,

Teaching physics in a physiologically meaningful
manner, Phys. Rev. ST Phys. Educ. Res. 6, 020116
(2010).
[373] J. L. Campbell, A quantitative comparison of the composition and linguistic properties of introductory physics
textbooks, Ph.D. thesis, University of Arkansas, 2008
(unpublished).
[374] Cognitive Load Theory, edited by J. L. Plass, R. Moreno,
and R. Brunken (Cambridge University Press, New York,
2010).
[375] J. Sweller, in The Psychology of Learning, and Motivation: Cognition in Education, edited by J. P. Mestre and
B. H. Ross (Elsevier, San Diego, 2011).
[376] I. Halloun and D. Hestenes, The initial knowledge state of
college physics students, Am. J. Phys. 53, 1043 (1985).
[377] See a list at http://www.ncsu.edu/per/TestInfo.html.
[378] P. V. Engelhardt, in Reviews in PER Vol. 2: Getting
Started in Physics Education Research, edited by C.
Henderson and K. A. Harper (American Association of
Physics Teachers, College Park, MD, 2009), http://www
.percentral.org/document/ServeFile.cfm?ID=8807, retrieved August 10, 2010.
[379] D. Treagust, Evaluating students misconceptions by
means of diagnostic multiple choice items, Res. Sci.
Educ. 16, 199 (1986).
[380] R. S. Lindell, E. Peak, and T. M. Foster, Are they all
created equal? A comparison of different concept inventory development methodologies, AIP Conf. Proc. 883,
14 (2007).
[381] L. Ding and R. Beichner, Approaches to data analysis of
multiple-choice questions, Phys. Rev. ST Phys. Educ.
Res. 5, 020103 (2009).
[382] L. Ding, N. W. Reay, A. Lee, and L. Bao, Effects of
testing conditions on conceptual survey results, Phys.
[383] V. P. Coletta, J. A. Phillips, and J. J. Steinert, Interpreting
force concept inventory scores: Normalized gain and
SAT scores, Phys. Rev. ST Phys. Educ. Res. 3, 010106
(2007).
[384] V. P. Coletta and J. A. Phillips, Interpreting FCI scores:
Normalized gain, preinstruction scores, and scientific
reasoning ability, Am. J. Phys. 73, 1172 (2005).
[385] D. E. Meltzer, The relationship between mathematics
preparation and conceptual learning gains in physics: A
possible hidden variable in diagnostic pretest scores,
Am. J. Phys. 70, 1259 (2002).
[386] E. Kuo, M. M. Hull, A. Gupta, and A. Elby, How students
blend conceptual and formal mathematical reasoning in
solving physics problems, Sci. Educ. 97, 32 (2013).
[387] R. K. Thornton, D. Kuhl, K. Cummings, and J. Marx,
Comparing the force and motion conceptual evaluation
and the force concept inventory, Phys. Rev. ST Phys.
Educ. Res. 5, 010105 (2009).
[388] L. Bao, T. Cai, K. Koenig, K. Fang, J. Han, J. Wang, Q.
Liu, L. Ding, L. Cui, Y. Luo, Y. Luo, Y. Wang, L. Li, and
N. Wu, Learning and scientific reasoning, Science 323,
586 (2009).
[389] L. Bao, K. Fang, T. Cai, J. Wang, L. Yang, L. Cui, J. Han,
L. Ding, and Y. Luo, Learning of content knowledge and
[390]
[391]
[392]
[393]
[394]
[395]
[396]
[397]
[398]
[399]
[400]
[401]
[402]
[403]
[404]
[405]
[406]
[407]
020119-54
development of scientific reasoning ability: A cross

culture comparison, Am. J. Phys. 77, 1118 (2009).
L. E. Kost, S. J. Pollock, and N. D. Finkelstein, Characterizing the gender gap in introductory physics, Phys. Rev.
ST Phys. Educ. Res. 5, 010101 (2009).
M. Lorenzo, C. H. Crouch, and E. Mazur, Reducing the
gender gap in the physics classroom, Am. J. Phys. 74, 118
(2006).
S. J. Pollock, N. D. Finkelstein, and L. E. Kost, Reducing
the gender gap in the physics classroom: How sufficient is
interactive engagement?, Phys. Rev. ST Phys. Educ. Res.
3, 010107 (2007).
G. Kortemeyer, Gender differences in the use of an
online homework system in an introductory physics
course, Phys. Rev. ST Phys. Educ. Res. 5, 010107 (2009).
C. T. Richardson and B. W. OShea, Assessing gender
differences in response system questions for an introductory physics course, Am. J. Phys. 81, 231 (2009).
H. T. Hudson and C. K. Hudson, Suggestions on the
construction of multiple-choice tests, Am. J. Phys. 49,
838 (1981).
M. Scott, T. Stelzer, and G. Gladding, Evaluating multiplechoice exams in large introductory physics courses, Phys.
M. M. Hull, E Kuo, A. Gupta, and A. Elby, Problemsolving rubrics revisited: Attending to blending of informal conceptual and formal mathematical reasoning,
J. D. Marx and K. Cummings, Development of a survey
instrument to gauge students problem-solving ability,
AIP Conf. Proc. 1289, 221 (2010).
L. Bao and E. F. Redish, Model analysis: Representing
and assessing the dynamics of student learning, Phys.
D. E. Pritchard, Y.-J. Lee, and L. Bao, Mathematical
learning models that depend on prior knowledge and
instructional strategies, Phys. Rev. ST Phys. Educ. Res. 4,
010109 (2008).
A. F. Heckler and E. C. Sayre, What happens between
pre- and post-tests: Multiple measurements of student
understanding during an introductory physics course,
Am. J. Phys. 78, 768 (2010).
E. C. Sayre and A. F. Heckler, Peaks and decays of
student knowledge in an introductory E&M course, Phys.
See http://www.teststandards.org.
American Educational Research Association, American
Psychological Association, and National Council on
Measurement in Education, Standards for Educational
and Psychological Testing (American Educational
Research Association, Washington, DC, 1999).
P. Heller and D. Huffman, Interpreting the Force Concept
Inventory: A reply to Hestenes and Halloun, Phys. Teach.
33, 503 (1995).
D. Hestenes and I. Halloun, Interpreting the Force
Concept Inventory: A response to March 1995 critique
by Huffman and Heller, Phys. Teach. 33, 502 (1995).
D. Huffman and P. Heller, What does the Force Concept
Inventory actually measure?, Phys. Teach. 33, 138
(1995).
[408] S. Ramlo, Validity and reliability of the force and

motion conceptual evaluation, Am. J. Phys. 76, 882
(2008).
[409] C. Singh and D. Rosengrant, Multiple-choice test of
energy and momentum concepts, Am. J. Phys. 71, 607
(2003).
[410] P. V. Engelhardt and R. J. Beichner, Students understanding of direct current resistive electrical circuits,
Am. J. Phys. 72, 98 (2004).
[411] L. Ding, R. Chabay, B. Sherwood, and R. Beichner,
Evaluating an electricity and magnetism assessment tool:
Brief electricity and magnetism assessment, Phys. Rev.
ST Phys. Educ. Res. 2, 010105 (2006).
[412] P. Nieminen, A. Savinainen, and J. Viiri, Force concept
inventory-based multiple-choice test for investigating
students representational consistency, Phys. Rev. ST
Phys. Educ. Res. 6, 020109 (2010).
[413] S. B. McKagan, K. K. Perkins, and C. E. Wieman, Design
and validation of the Quantum Mechanics Conceptual
Survey, Phys. Rev. ST Phys. Educ. Res. 6, 020121
(2010).
[414] S. V. Chasteen, R. E. Pepper, M. D. Caballero, S. J.
Pollock, and K. K. Perkins, Colorado Upper-Division
Electrostatics diagnostic: A conceptual assessment for
the junior level, Phys. Rev. ST Phys. Educ. Res. 8,
020108 (2012).
[415] R. Rosenblatt and A. F. Heckler, Systematic study of
student understanding of the relationships between the
direction of force, velocity, and acceleration in one
dimension, Phys. Rev. ST Phys. Educ. Res. 7, 020112
(2011).
[416] A. E. Lawson, Development and validation of the classroom test of formal reasoning, J. Res. Sci. Teach. 15, 11
(1978).
[417] L. Bao, Theoretical comparisons of average normalized
gain calculations, Am. J. Phys. 74, 917 (2006).
[418] J. D. Marx and K. Cummings, Normalized change, Am. J.
Phys. 75, 87 (2007).
[419] G. A. Morris, L. Branum-Martin, N. Harshman, S. D.
Baker, E. Mazur, S. Dutta, T. Mzoughi, and V. McCauley,
Testing the test: Item response curves and test quality,
Am. J. Phys. 74, 449 (2006).
[420] J. Wang and L. Bao, Analyzing force concept inventory
with item response theory, Am. J. Phys. 78, 1064 (2010).
[421] T. I. Smith and M. C. Wittmann, Applying a resources
framework to analysis of the Force and Motion Conceptual Evaluation, Phys. Rev. ST Phys. Educ. Res. 4,
020101 (2008).
[422] J. Stewart, M. Miller, C. Audo, and G. Stewart, Using
cluster analysis to identify patterns in students responses
to contextually different conceptual problems, Phys. Rev.
ST Phys. Educ. Res. 8, 020112 (2012).
[423] M. Planinic, L. Ivanjek, and A. Susac, Rasch model based
analysis of the Force Concept Inventory, Phys. Rev. ST
Phys. Educ. Res. 6, 010103 (2010).
[424] L. Bao and E. F. Redish, Concentration analysis: A
quantitative assessment of student states, Am. J. Phys.
69, S45 (2001).
[425] C. Henderson, Common concerns about the Force
Concept Inventory, Phys. Teach. 40, 542 (2002).
[426] S. Bonham, Reliability, compliance, and security in webbased course assessments, Phys. Rev. ST Phys. Educ.
Res. 4, 010106 (2008).
[427] J. Stewart, H. Griffin, and G. Stewart, Context sensitivity
in the force concept inventory, Phys. Rev. ST Phys. Educ.
Res. 3, 010102 (2007).
[428] B. W. Frank and R. E. Scherr, Interactional processes for
stabilizing conceptual coherences in physics, Phys. Rev.
ST Phys. Educ. Res. 8, 020101 (2012).
[429] A. B. Champagne and L. E. Klopfer, A causal model of
students achievement in a college physics course, J. Res.
Sci. Teach. 19, 299 (1982).
[430] A. B. Champagne, L. E. Klopfer, and J. H. Anderson,
Factors influencing the learning of classical mechanics,
Am. J. Phys. 48, 1074 (1980).
[431] J. Docktor and K. Heller, Gender differences in both force
concept inventory and introductory physics performance,
AIP Conf. Proc. 1064, 15 (2008).
[432] Z. Hazari, R. H. Tai, and P. M. Sadler, Gender differences
in introductory physics performance: The influence of
high school physics preparation and affective factors, Sci.
Educ. 91, 847 (2007).
[433] P. M. Sadler and R. H. Tai, Success in introductory
college physics: The role of high school preparation,
Sci. Educ. 85, 111 (2001).
[434] R. H. Tai and P. M. Sadler, Gender differences in introductory undergraduate physics performance: University
physics versus college physics in the USA, Int. J. Sci.
Educ. 23, 1017 (2001).
[435] L. E. Kost-Smith, S. J. Pollock, and N. D. Finkelstein,
Gender disparities in second-semester college physics:
The incremental effects of a smog of bias, Phys. Rev.
ST Phys. Educ. Res. 6, 020112 (2010).
[436] A. Miyake, L. E. Kost-Smith, N. D. Finkelstein, S. J.
Pollock, G. L. Cohen, and T. A. Ito, Reducing the gender
achievement gap in college science: A classroom study of
values affirmation, Science 330, 1234 (2010).
[437] W. Fakcharoenphol, E. Potter, and T. Stelzer, What
students learn when studying physics practice exam
(2011).
[438] D. J. Palazzo, Y.-L. Lee, R. Warnakulasooriya, and D. E.
Pritchard, Patterns, correlates, and reduction of homework copying, Phys. Rev. ST Phys. Educ. Res. 6, 010104
(2010).
[439] K. Cummings and J. D. Marx, Beta-test data on an
assessment of textbook problem solving ability: An
argument for right/wrong grading?, AIP Conf. Proc.
1289, 113 (2010).
[440] J. Stewart and S. Ballard, Effect of written presentation on
performance in introductory physics, Phys. Rev. ST Phys.
Educ. Res. 6, 020120 (2010).
[441] M. A. McDaniel, H. L. Roediger III, and K. B.
McDermott, Generalizing test-enhanced learning from
the laboratory to the classroom, Psychon. Bull. Rev.
14, 200 (2007).
[442] For a discussion of the relationship between disciplinebased education research and other research areas such as
educational psychology and cognitive science, refer to the
introduction of Ref. [2].
020119-55
[443] M. L. Gick and K. J. Holyoak, Schema induction and

analogical transfer, Cogn. Psychol. 15, 1 (1983).
[444] S. K. Reed, in Transfer on Trial: Intelligence, Cognition,
and Instruction, edited by D. K. Detterman and
R. J. Sternberg (Ablex, Norwood, NJ, 1993),
pp. 3967.
[445] A. Feil and J. Mestre, Change blindness as a means of
studying expertise in physics, J. Learn. Sci. 19, 480
(2010).
[446] J. Mestre, Is transfer ubiquitous or rare? New paradigms
for studying transfer, AIP Conf. Proc. 790, 3 (2005).
[447] R. Dufresne, J. P. Mestre, T. Thaden-Koch, W. Gerace,
and W. Leonard, in Transfer of Learning from a Modern
Multidisciplinary Perspective, edited by J. P. Mestre
(Information Age Publishing, Greenwich, CT, 2005),
pp. 155215.
[448] C. Singh, Transfer of learning in quantum mechanics, AIP
Conf. Proc. 790, 23 (2005).
[449] T. J. Nokes-Malach and J. P. Mestre, Toward a model
of transfer as sense-making, Educ. Psychol. 48, 184
(2013).
[450] N. S. Rebello, D. A. Zollman, A. R. Allbaugh, P. V.
Engelhardt, K. E. Gray, Z. Hrepic, and S. F. Itza-Ortiz,
in Transfer of Learning from a Modern Multidisciplinary
Perspective, edited by J. P. Mestre (Information Age
Publishing, Greenwich, CT, 2005).
[451] J. Mestre, Transfer of learning: Issues and research
agenda, National Science Foundation Report
No. NSF03-212, 2003, www.nsf.gov/pubs/2003/
nsf03212/nsf03212.pdf, retrieved October 4, 2010.
[452] Special issue on transfer strand [J. Learn. Sci. 15, Issue
No. 4, 431 (2006)].
[453] The Cambridge Handbook of Expertise and Expert
Performance, edited by K. A. Ericsson, N. Charness, P.
Feltovich, and R. Hoffman (Cambridge University Press,
Cambridge, England, 2006).
[454] Transfer of Learning from a Modern Multidisciplinary
Perspective, edited by J. P. Mestre (Information Age
Publishing, Greenwich, CT, 2005).
[455] J. H. Flavell, Metacognition and cognitive monitoring:
A new area of cognitive-developmental inquiry, Am.
Psychol. 34, 906 (1979).
[456] G. Taasoobshirazi and J. Farley, Construct validation of
the Physics Metacognitive Inventory, Int. J. Sci. Educ. 35,
447 (2013).
[457] D. L. Medin, Concepts and conceptual structure, Am.
Psychol. 44, 1469 (1989).
[458] D. L. Medin and L. J. Rips, in The Cambridge Handbook
of Thinking and Reasoning, edited by K. J. Holyoak and
R. G. Morrison (Cambridge University Press, New York,
2005), pp. 3772.
[459] B. L. Whorf, in Language, Thought, and Reality: Selected
Writings of Benjamin Lee Whorf, edited by J. B. Carroll
(The MIT Press, Cambridge, MA, 1956).
[460] C. R. Sutton, Figuring out a scientific understanding,
J. Res. Sci. Teach. 30, 1215 (1993).
[461] D. T. Brookes and E. Etkina, Using conceptual metaphor
and functional grammar to explore how language used in
physics affects student learning, Phys. Rev. ST Phys.
Educ. Res. 3, 010105 (2007).
[462] D. T. Brookes and E. Etkina, Force, ontology, and

language, Phys. Rev. ST Phys. Educ. Res. 5, 010110
(2009).
[463] N. S. Podolefsky and N. D. Finkelstein, Use of analogy in
learning physics: The role of representations, Phys. Rev.
ST Phys. Educ. Res. 2, 020101 (2006).
[464] N. S. Podolefsky and N. D. Finkelstein, Analogical scaffolding and the learning of abstract ideas in physics: An
example from electromagnetic waves, Phys. Rev. ST
Phys. Educ. Res. 3, 010109 (2007).
[465] N. S. Podolefsky and N. D. Finkelstein, Analogical scaffolding and the learning of abstract ideas in physics:
Empirical studies, Phys. Rev. ST Phys. Educ. Res. 3,
020104 (2007).
[466] A. A. diSessa, An interactional analysis of clinical interviewing, Cognit. Instr. 25, 523 (2007).
[467] K. A. Ericsson and H. A. Simon, Protocol Analysis:
Verbal Reports as Data (revised edition) (Bradford
Books, Cambridge, MA, 1993).
[468] V. R. Lee, R. S. Russ, and B. Sherin, in Proceedings
of the 30th Annual Meeting of the Cognitive
Science Society, Austin, TX, 2008, edited by V. Sloutsky,
B. Love, and K. McRae (Cognitive Science Society,
2008), pp. 17231728, http://csjarchive.cogsci.rpi.edu/
Proceedings/2008/pdfs/p1723.pdf.
[469] R. E. Nisbett and T. D. Wilson, Telling more than we can
know: Verbal reports on mental processes, Psychol. Rev.
84, 231 (1977).
[470] M. T. H. Chi, N. De Leeuw, M.-H. Chiu, and C.
Lavancher, Eliciting self-explanations improves understanding, Cogn. Sci. 18, 439 (1994).
[471] L. C. McDermott and P. S. Shaffer, Tutorials in
Introductory Physics (Prentice-Hall, Upper Saddle River,
NJ, 1992).
[472] J. Confrey, in The Cambridge Handbook of the
Learning Sciences, edited by R. K. Sawyer (Cambridge
University Press, Cambridge, England, 2006),
pp. 135152.
[473] T. Lamberg and J. A. Middleton, Design research perspectives on transitioning from individual microgenetic
interviews to a whole-class teaching experiment, Educ.
Res. 38, 233 (2009).
[474] M. T. H. Chi, in Advances in Instructional Psychology:
Educational Design and Cognitive Science, edited by R.
Glaser (Lawrence Erlbaum Assoc., Mahwah, NJ, 2000),
Vol. 5, pp. 161238.
[475] A. B. Hoffman and B. Rehder, The costs of supervised
classification: The effect of learning task on conceptual
flexibility, J. Exp. Psychol. 139, 319 (2010).
[476] K. Rayner, Eye movements in reading and information
processing: 20 years of research, Psychol. Bull. 124, 372
(1998).
[477] D. J. Simons and D. T. Levin, Failure to detect changes to
people during a real-world interaction, Psychon. Bull.
Rev. 5, 644 (1998).
[478] C. Singh, Categorization of problems to assess and
improve proficiency as teachers and learners, Am. J.
Phys. 77, 73 (2009).
[479] G. Bateson, in Steps to an Ecology of Mind: Collected
Essays in Anthropology, Psychiatry, Evolution, and
020119-56
[480]
[481]
[482]
[483]
[484]
[485]
[486]
[487]
[488]
[489]
[490]
[491]
[492]
[493]
[494]
Epistemology, edited by G. Bateson (Ballantine,

New York, 1972), pp. 177193.
E. Goffman, Frame Analysis: An Essay on the
Organization of Experience (Northeastern University
Press, Boston, MA, 1986).
G. L. MacLachlan and I. Reid, Framing and Interpretation (Melbourne University Press, Portland, OR, 1994).
D. Tannen, Framing in Discourse (Oxford University
Press, New York, 1993).
E. F. Redish, A theoretical framework for physics
education research: Modeling student thinking, in
Proceedings of the International School of Physics
Enrico Fermi, Course CLVI, Varenna, Italy, 2003,
edited by E. Redish, C. Tarsitani, and M. Vicentini (IOS
Press, Amsterdam, 2004).
B. H. Ross, Remindings and their effects in
learning a cognitive skill, Cogn. Psychol. 16, 371
(1984).
D. T. Brookes, B. H. Ross, and J. P. Mestre, Specificity,
transfer, and the development of expertise, Phys. Rev. ST
Phys. Educ. Res. 7, 010105 (2011).
J. P. Mestre, B. H. Ross, D. T. Brookes, A. D. Smith, and
T. J. Nokes, in Fostering Scientific Habits of Mind:
Pedagogical Knowledge and Best Practices in Science
Education, edited by I. M. Saleh and M. S. Khine
(Sense Publishers, Rotterdam, Netherlands, 2009),
pp. 145171.
A. M. Madsen, A. M. Larson, L. C. Loschky, and
N. S. Rebello, Differences in visual attention between
those who correctly and incorrectly answer physics
(2012).
T. Van Gog, F. Paas, and J. J. G. Van Merrinboer,
Uncovering expertise-related differences in troubleshooting performance: Combining eye movement data
and concurrent verbal protocol data, Appl. Cogn.
Psychol. 19, 205 (2005).
A. C. Graesser, S. Lu, B. A. Olde, E. Cooper-Pye, and S.
Whitten, Question asking and eye tracking during cognitive disequilibrium: Comprehending illustrated texts on
devices when the devices break down, Mem. Cogn. 33,
1235 (2005).
M. Hegarty, in Eye Movements and Visual Cognition,
edited by K. Rayner (Springer-Verlag, New York, 1992),
pp. 428443.
M. Kozhevnikov, M. A. Motes, and M. Hegarty, Spatial
visualization in physics problem solving, Cogn. Sci. 31,
549 (2007).
D. Rosengrant, C. Thomson, and T. Mzoughi, Comparing
experts and novices in solving electrical circuit problems
with the help of eye-tracking, AIP Conf. Proc. 1179, 249
(2009).
A. Mason and C. Singh, Assessing expertise in introductory physics using categorization task, Phys. Rev. ST
Phys. Educ. Res. 7, 020110 (2011).
R. K. Atkinson, S. J. Derry, A. Renkl, and D. W.
Wortham, Learning from examples: Instructional principles from the worked examples research, Rev. Educ. Res.
70, 181 (2000).
[495] J. Sweller, J. J. G. Van Merrinboer, and F. G. W. C. Paas,

Cognitive architecture and instructional design, Educ.
Psychol. Rev. 10, 251 (1998).
[496] F. Paas, Training strategies for attaining transfer of
problem-solving skill in statistics: A cognitive-load
approach, J. Educ. Psychol. 84, 429 (1992).
[497] F. G. W. C. Paas and J. J. G. Van Merrinboer, Variability
of worked examples and transfer of geometrical problemsolving skills: A cognitive-load approach, J. Educ.
Psychol. 86, 122 (1994).
[498] M. T. H. Chi and K. VanLehn, The content of physics
self-explanations, J. Learn. Sci. 1, 69 (1991).
[499] D. T. Brookes, The role of language in learning physics,
Ph.D. thesis (Rutgers, The State University of New Jersey,
2006 (unpublished).
[500] K. K. Perkins, W. K. Adams, S. J. Pollock, N. D. Finkelstein,
and C. E. Wieman, Correlating student beliefs with student
learning using the Colorado Learning Attitudes about
Science Survey, AIP Conf. Proc. 790, 61 (2005).
[501] D. Hammer, Epistemological beliefs in introductory
physics, Cognit. Instr. 12, 151 (1994).
[502] D. Hammer, Epistemological considerations in teaching
introductory physics, Sci. Educ. 79, 393 (1995).
[503] I. Halloun, Views about science, and physics achievement: The VASS story, AIP Conf. Proc. 399, 605 (1997).
[504] U. Wutchana and N. Emarat, Student effort expectations
and their learning in first-year introductory physics: A
case study in Thailand, Phys. Rev. ST Phys. Educ. Res. 7,
010111 (2011).
[505] D. Hammer and A. Elby, Tapping epistemological
resources for learning physics, J. Learn. Sci. 12, 53
(2003).
[506] E. F. Redish, J. M. Saul, and R. N. Steinberg, Student
expectations in introductory physics, Am. J. Phys. 66, 212
(1998).
[507] I. Halloun and D. Hestenes, Interpreting VASS dimensions and profiles for physics students, Sci. Educ. 7, 553
(1998).
[508] B. White, A. Elby, J. Frederiksen, and C. Schwarz, in
Proceedings of the American Education Research
Association, Montreal, 1999 (unpublished).
[509] W. K. Adams, K. K. Perkins, N. S. Podolefsky, M.
Dubson, N. D. Finkelstein, and C. E. Wieman, New
instrument for measuring student beliefs about physics
and learning physics: The Colorado Learning Attitudes
about Science Survey, Phys. Rev. ST Phys. Educ. Res. 2,
010101 (2006).
[510] A. Mason and C. Singh, Surveying graduate students
attitudes and approaches to problem solving, Phys. Rev.
ST Phys. Educ. Res. 6, 020124 (2010).
[511] E. Yerushalmi, C. Henderson, K. Heller, P. Heller, and V.
Kuo, Physics faculty beliefs and values about the teaching
and learning of problem solving. I: Mapping the common
core, Phys. Rev. ST Phys. Educ. Res. 3, 020109
(2007).
[512] C. Henderson and M. H. Dancy, Barriers to the use of
research-based instructional strategies: The influence of
both individual and situational characteristics, Phys. Rev.
ST Phys. Educ. Res. 3, 020102 (2007).
020119-57
[513] C. Henderson, E. Yerushalmi, V. H. Kuo, K. Heller, and P.

Heller, Physics faculty beliefs and values about the
teaching and learning of problem solving. II: Procedures
for measurement and analysis, Phys. Rev. ST Phys. Educ.
Res. 3, 020110 (2007).
[514] E. Yerushalmi, E. Cohen, K. Heller, P. Heller, and C.
Henderson, Instructors reasons for choosing problem
features in a calculus-based introductory physics course,
[515] C. Henderson, A. Beach, and N. Finkelstein, Facilitating
change in undergraduate STEM instructional practices:
An analytic review of the literature, J. Res. Sci. Teach. 48,
952 (2011).
[516] C. Henderson and M. H. Dancy, Physics faculty and
educational researchers: Divergent expectations as barriers to the diffusion of innovations, Am. J. Phys. 76, 79
(2008).
[517] C. Turpen and N. D. Finkelstein, Not all interactive
engagement is the same: Variations in physics professors
implementation of Peer Instruction, Phys. Rev. ST Phys.
Educ. Res. 5, 020101 (2009).
[518] C. Turpen and N. D. Finkelstein, The construction of
different classroom norms during Peer Instruction: Students perceive differences, Phys. Rev. ST Phys. Educ.
Res. 6, 020123 (2010).
[519] R. M. Goertzen, R. E. Scherr, and A. Elby, Accounting for
tutorial teaching assistants buy-in to reform instruction,
[520] R. M. Goertzen, R. E. Scherr, and A. Elby, Tutorial
teaching assistants in the classroom: Similar teaching
behaviors are supported by varied beliefs about teaching
and learning, Phys. Rev. ST Phys. Educ. Res. 6, 010105
(2010).
[521] K. M. Koenig, R. J. Endorf, and G. A. Braun, Effectiveness of different tutorial recitation teaching methods and
its implications for TA training, Phys. Rev. ST Phys.
Educ. Res. 3, 010104 (2007).
[522] D. Hammer, Two approaches to learning physics, Phys.
Teach. 27, 664 (1989).
[523] M. Dancy and C. Henderson, Framework for articulating
instructional practices and conceptions, Phys. Rev. ST
Phys. Educ. Res. 3, 010103 (2007).
[524] G. Kortemeyer, Correlations between student discussion
behavior, attitudes, and learning, Phys. Rev. ST Phys.
Educ. Res. 3, 010101 (2007).
[525] B. K. Hofer, Personal epistemology research: Implications for learning and teaching, J. Educ. Psychol. Rev. 13,
353 (2001).
[526] B. K. Hofer and P. R. Pintrich, The development of
epistemological theories: Beliefs about knowledge and
[527]
[528]
[529]
[530]
[531]
[532]
[533]
[534]
[535]
[536]
[537]
[538]
[539]
020119-58
knowing and their relation to learning, Rev. Educ. Res.

67, 88 (1997).
A. Elby, Another reason that physics students learn by
rote, Am. J. Phys. 67, S52 (1999).
K. E. Gray, W. K. Adams, C. E. Wieman, and K. K.
Perkins, Students know what physicists believe, but they
dont agree: A study using the CLASS survey, Phys. Rev.
ST Phys. Educ. Res. 4, 020106 (2008).
E. Gire, B. Jones, and E. Price, Characterizing the
epistemological development of physics majors, Phys.
B. A. Lindsey, L. Hsu, H. Sadaghiani, J. W. Taylor,
K. Cummings, Positive attitudinal shifts with the
Physics by Inquiry curriculum across multiple implementations, Phys. Rev. ST Phys. Educ. Res. 8, 010102
(2012).
M. Milner-Bolotin, T. Antimirova, A. Noack, and
A. Petrov, Attitudes about science and conceptual
physics learning in university introductory physics
courses, Phys. Rev. ST Phys. Educ. Res. 7, 020107
(2011).
J. D. H. Gaffney, A. L. Housley Gaffney, and R. J.
Beichner, Do they see it coming? Using expectancy
violation to gauge the success of pedagogical reforms,
M. Bodin and M. Winberg, Role of beliefs and emotions
in numerical problem solving in university physics
education, Phys. Rev. ST Phys. Educ. Res. 8, 010108
(2012).
L. Lising and A. Elby, The impact of epistemology on
learning: A case study from introductory physics, Am. J.
Phys. 73, 372 (2005).
B. T. Spike and N. D. Finkelstein, Examining the beliefs
and practice of teaching assistants: Two case studies, AIP
Conf. Proc. 1289, 309 (2010).
S.-Y. Lin, C. Henderson, W. Mamudi, C. Singh, and
E. Yerushalmi, Teaching assistants beliefs regarding
example solutions in introductory physics, Phys. Rev.
ST Phys. Educ. Res. 9, 010120 (2013).
G. DeBeck, S. Settelmeyer, S. Li, and D. Demaree, TA
beliefs in a SCALE-UP style classroom, AIP Conf. Proc.
1289, 121 (2010).
E. A. West, C. A. Paul, C. Webb, and W. H. Potter,
Variation of instructor-student interactions in an introductory interactive physics course, Phys. Rev. ST Phys.
Educ. Res. 9, 010109 (2013).
M. J. Volkmann and M. Zgagacz, Learning to teach
physics through inquiry: The lived experience of a
graduate teaching assistant, J. Res. Sci. Teach. 41, 584
(2004).

Synthesis of Discipline-Based Education Research in Physics

Uploaded by

Copyright:

Available Formats

Synthesis of Discipline-Based Education Research in Physics

Uploaded by

Document Information

Original Description:

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Synthesis of Discipline-Based Education Research in Physics

Uploaded by

Copyright:

Available Formats

PHYSICAL REVIEW SPECIAL TOPICS - PHYSICS EDUCATION RESEARCH 10, 020119 (2014)

Synthesis of discipline-based education research in physics

Department of Physics, University of Wisconsin-La Crosse, 1725 State Street,

PACS numbers: 01.40.Ha, 01.40.Fk, 01.40.gb

Published by the American Physical Society under the terms of

Published by the American Physical Society

JENNIFER L. DOCKTOR AND JOS P. MESTRE

PHYS. REV. ST PHYS. EDUC. RES 10, 020119 (2014)

although not exhaustive; whenever multiple articles were

common student misconceptions [4,5], and there were

B. Possible uses of this synthesis

What learning difficulties do students possess that are

PHYS. REV. ST PHYS. EDUC. RES 10, 020119 (2014)

architectures [1823], and some have combined different

from these and multiple similar observations construct the

3. Developing and evaluating instructional strategies to

2. Knowledge in pieces or resources view

JENNIFER L. DOCKTOR AND JOS P. MESTRE

PHYS. REV. ST PHYS. EDUC. RES 10, 020119 (2014)

students knowledge consists of smaller grain-size pieces

ontological categories and that many of the difficulties

3. Ontological categories view

A third perspective of novice knowledge and conceptual

C. Methodology (data collection or sources

PHYS. REV. ST PHYS. EDUC. RES 10, 020119 (2014)

large amounts of data that are difficult to analyze and

between the ontological categories view and the other

JENNIFER L. DOCKTOR AND JOS P. MESTRE

PHYS. REV. ST PHYS. EDUC. RES 10, 020119 (2014)

E. Strengths and limitations of conceptual

In addition to research on conceptual understanding,

PHYS. REV. ST PHYS. EDUC. RES 10, 020119 (2014)

(Sec. VI). For additional reviews of research on physics

solving [100,101], as well as whether problems use

JENNIFER L. DOCKTOR AND JOS P. MESTRE

PHYS. REV. ST PHYS. EDUC. RES 10, 020119 (2014)

representing the problem, a solver engages in a search

emphasize the role of specific example problems for

2. Problem solving by analogy

3. Resources model and epistemic games

PHYS. REV. ST PHYS. EDUC. RES 10, 020119 (2014)

could include representations (such as pictures or physics

JENNIFER L. DOCKTOR AND JOS P. MESTRE

PHYS. REV. ST PHYS. EDUC. RES 10, 020119 (2014)

students to correct mistakes on assessments improved

PHYS. REV. ST PHYS. EDUC. RES 10, 020119 (2014)

They concluded that these descriptions markedly

Vectors. Studies have documented several student

JENNIFER L. DOCKTOR AND JOS P. MESTRE

PHYS. REV. ST PHYS. EDUC. RES 10, 020119 (2014)

unfamiliar ideas, suggesting a plasticity continuum

Qualitative approaches. Some curricula emphasize

PHYS. REV. ST PHYS. EDUC. RES 10, 020119 (2014)

for improvements (or decrements) in problem-solving

most prominent gaps in research on problem solving

F. Areas for future study

IV. CURRICULUM AND INSTRUCTION

JENNIFER L. DOCKTOR AND JOS P. MESTRE