DATA COLLECTION

Pitfalls
It is often difficult to keep the study focused on a few important variables and hypotheses. New and interesting possibilities frequently appear. Some researchers may use a shotgun approach, collecting everything that might be useful. However, the more variables that you have, the more difficult the research process becomes. It is also more likely that you will confuse major and minor variables. More is not better unless you are planning on a series of research initiatives based on one larger pool of data with different variables being used at different times. Being selective in what we collect and what we ask is one of the most difficult of all research tasks. Be vigilant and remind yourself regularly just what your study is about and what is the heart of the matter.

Necessities

Your population and a sampling frame must be clearly identified and both must make sense. Each variable should be clearly identified and their values should be operationally defined. A code book should be available to guide data collection.
Literature Review

You must be able to identify and characterize your research methods. Once you know the proper names, you should do a reasonable literature review on the methods that you will use to learn more about their assets and liabilities. Resources may include research methods textbooks, articles on the method, and articles on various research topics that have used the method. You are responsible for being familiar with the major strengths and weaknesses of your methods. You may have to defend them with an editor or those who review your work.

Transparency
The data collection process must be full revealed so you will need to think through the process on a step-by-step basis. It's a good idea to list each separate step in the data collection process as if you were going to flow chart it. Review your list to insure that it is as logical and complete as possible. Test your list by applying it to several representative cases.

Problems
Identify and list snags or problems encountered or likely to be encountered. For each obstacle identified, find one or more ways to compensate or overcome the problem. There are many sources of error in data collection. Here are a few: Transient personal factors such as health, fatigue, motivation Situational factors such as relations with colleagues, distractions Instrumentation problems such as lack of clarity in an interview schedule, troublesome arrangement/organization on a questionnaire, or response factors such as respondents selecting "no" rather than "yes" because "yes" leads to more questions Analysis factors such as errors in scoring, tabulation, or the use of an inappropriate statistical test. Here is an example of a data collection problem that was overcome. A questionnaire with a low response rate on a pilot test was redesigned to include an endorsement cover letter, reduced length, and a better benefit/rationale statement. If problems cannot be overcome, it may be necessary to change the research design or delete a troublesome variable. You are responsible for convincing the reader that your data collection method was appropriate and free from error.

Quality Control
It is important to record problems encountered and steps taken in your research diary so that you can discover what you did and why later. This is especially important if you change or amend an operational definition and wish to be consistent in the future. Reliability is of particular concern with data collection, especially if others are involved. Check your coding sheets of inappropriate values such as the fourteen year old widows found in an old census document. Reliability testing with another is an important and needed step.

DATA COLLECTION METHODS

Data collection is a term used to describe a process of preparing and collecting data - for example as part of a process improvement or similar project. The purpose of data collection is to obtain information to keep on record, to make decisions about important issues, to pass information on to others. Primarily, [1] data is collected to provide information regarding a specific topic. Data collection usually takes place early on in an improvement project, and is often formalised through [2] a data collection plan which often contains the following activity. 1. Pre collection activity Agree goals, target data, definitions, methods 2. Collection data collection 3. Present Findings usually involves some form of sorting
[3]

analysis and/or presentation.

Prior to any data collection, pre-collection activity is one of the most crucial steps in the process. It is often discovered too late that the value of their interview information is discounted as a consequence of poor [4] sampling of both questions and informants and poor elicitation techniques. After pre-collection activity is fully completed, data collection in the field, whether by interviewing or other methods, can be carried out in a structured, systematic and scientific way. A formal data collection process is necessary as it ensures that data gathered is both defined and [5] accurate and that subsequent decisions based on arguments embodied in the findings are valid. The process provides both a baseline from which to measure from and in certain cases a target on what to improve. Other main types of collection include census, sample survey, and administrative by-product and each with their respective advantages and disadvantages. A census refers to data collection about everyone or everything in a group or population and has advantages, such as accuracy and detail and disadvantages, such as cost and time. A sample survey is a data collection method that includes only part of the total population and has advantages, such as cost and time and disadvantages, such as accuracy and detail. Administrative by-product data is collected as a byproduct of an organization's day-to-day operations and has advantages, such as accuracy, time simplicity and disadvantages, such as no flexibility and lack of control.

The methods involved in survey data collection are any of a number of ways in which data can be collected for a statistical survey. These are methods that are used to collect information from a sample of individuals in a systematic way. [edit]Modes

of data collection

There are several ways of administering a survey. And within a survey, different methods could be used for different parts. For example, interviewer administration could be used for some items but self-

administration for sensitive topics. The choice between administration modes is influenced by several factors, including 1) costs, 2) coverage of the target population, 3) flexibility of asking questions, 4) respondents willingness to participate and 5) response accuracy. Different methods create mode effects that change how respondents answer. The most common modes [1] of administration are listed : [edit]Telephone ADVANTAGES use of interviewers encourages sample persons to respond, leading to higher response rates. interviewers can increase comprehension of questions by answering respondents' questions. fairly cost efficient, depending on local call charge structure good for large national (or international) sampling frames
[2]

DISADVANTAGES
some potential for interviewer bias (e.g. some people may be more willing to discuss a sensitive issue with a female interviewer than with a male one) cannot be used for non-audio information (graphics, demonstrations, taste/smell samples) unreliable for consumer surveys in rural areas where telephone density is low
[3]

three types: traditional telephone interviews computer assisted telephone dialing computer assisted telephone interviewing (CATI)

[edit]Mail the questionnaire may be handed to the respondents or mailed to them, but in all cases they are returned to the researcher via mail.

Advantage cost is very low, since bulk postage is cheap in most countries

respondents can answer at their own convenience (allowing them to break up long surveys; also useful if they need to check records to answer a question) no interviewer bias introduced large amount of information can be obtained: some mail surveys are as long as 50 pages.

DISAD long time delays, often several months, before the surveys are returned and statistical analysis can begin not suitable for issues that may require clarification

Response rates can be improved by using mail panels members of the panel have agreed to participate panels can be used in longitudinal designs where the same respondents are surveyed several times [edit]Online

surveys

can use web or e-mail. Web is preferred over e-mail because interactive HTML forms can be used

ADV often inexpensive to administer very fast results easy to modify response rates can be improved by using Online panels - members of the panel have agreed to participate data sets created in real time used in large scale industries.

DISAD honesty of responses can be an issue if not password-protected, easy to manipulate by completing multiple times to skew results data creation, manipulation and reporting can be automated and/or easily exported into a format that can be read by PSPP, DAP or other statistical analysis software some are incentive based (such as Survey Vault or YouGov) may skew sample towards a younger demographic compared with CATI often difficult to determine/control selection probabilities, hindering quantitative analysis of data

Personal mall intercept survey

shoppers at malls are intercepted - they are either interviewed on the spot, taken to a room and interviewed, or taken to a room and given a self-administered questionnaire socially acceptable - people feel that a mall is a more appropriate place to do research than their home potential for interviewer bias fast easy to manipulate by completing multiple times to skew results

Hybrid Methodology
Researchers can combine several above methods for the data collection. For example, researchers can invite shoppers at malls, and send willing participants questionnaires by emails.

New Development
With the introduction of computers to the survey process, survey mode now includes combinations of [4] [5] different approaches or mixed-mode designs. . Some of the most common methods are : Computer-assisted personal interviewing (CAPI): The computer displays the questions on screen, the interviewer reads them to the respondent, and then enters the respondent's answers. Audio computer-assisted self-interviewing (audio CASI): The respondent operates the computer, the computer displays the question on the screen and plays recordings of the questions to the respondents, who then enters his/her answers. Computer-assisted telephone interviewing (CATI) Interactive voice response (IVR): The computer plays recordings of the questions to respondents over the telephone, who then respond by using the keypad of the telephone or speaking their answers aloud. Web surveys: The computer administers the questions online.

Collecting Primary Data Survey Based Questionnaires Interviews Observations

Postal response rate with reduce Use of a Pilot Study 20-30 minutes maximum for completion Diary - a self administering questionnaire Substitute for observation Interviews: Structured, Semi-structured and open

Particularly useful for gathering qualitative data Face to face and telephone Highly structured questions are good for eliciting quantitative data Open or unstructured interviews allow access to unpredictable information Interviews: Structured, Semi-structured and open Semi-structured interviews provide a hybrid approach. Good for dealing with domain experts. Need for preparation. Standardised Scales and Tests Use of existing and verified tests Usability tests standard usability scale Sense of presence tests SOPI Observation Participant versus non-participant Qualitative versus quantitative Know what you are looking at Method of recording Observation Direct observation Be aware of the Hawthorne effect Video recording Software logging Interactive observation Wizard of Oz Verbal protocols Interacting with Participants Approach people with courtesy Identify yourself, your intent, and what you are looking for Offer to compensate participants Describe how you will use the information and why it is valuable Get permission to use the information and any photographs of video recordings that you take Interacting with Participants Keep all the information that you gather confidential Let people know that they can decline to answer questions or stop participating at any time Maintain a non judgemental, relaxed and enjoyable atmosphere Finally, ensure that all studies are approved by the Universitys Ethics Committee Some examples Talk aloud protocol used during the study of virtual environments (BENOGO) Some examples Study of the Jencks Landform - Scottish Gallery of Modern Art

 Selecting Appropriate Data Collection Methods Chapter 6

It is a capital mistake to theorize before one has data. -Sir Arthur Conan Doyle as Sherlock Holmes Data Collection Options Data collection possibilities are wide and varied with any one method of collection not inherently better than any other Each has pros and cons that must be weighed up in view of a rich and complex context The Data Collection Process All methods of collection require rigorous and systematic design and execution that includes thorough planning well considered development effective piloting weighed modification deliberate implementation and execution appropriate management and analysis Surveys Surveying involves gathering information from individuals using a questionnaire Surveys can reach a large number of respondents generate standardized, quantifiable, empirical data - as well as some qualitative data and offer confidentiality / anonymity Designing survey instruments capable of generating credible data, however, can be difficult Survey Types Surveys can be descriptive or explanatory involve entire populations or samples of populations capture a moment or map trends can be administered in a number of ways Survey Construction Survey construction involves formulating questions and response categories writing up background information and instruction working through organization and length determining layout and design Interviewing Interviewing involves asking respondents a series of open-ended questions Interviews can generate both standardized quantifiable data, and more in-depth qualitative data However, the complexities of people and the complexities of communication can create many opportunities for miscommunication and misinterpretation Interview Types

Interviews can range from formal to informal structured to unstructured can be one on one or involve groups Conducting Interviews When conducting your interviews you will need to question, prompt, and probe in ways that help you gather rich data actively listen and make sense of what is being said manage the overall process Observation Observation relies on the researchers ability to gather data though their senses - and allows researchers to document actual behaviour rather than responses related to behaviour However, the observed can act differently when surveilled, and observations can be tainted by a researchers worldview Observation Types Observation can range from non-participant to participant candid to covert from structured to unstructured The Observation Process The observation process is sometimes treated casually, but is a method that needs to be treated as rigorously as any other The process should include planning, observing, recording, reflecting, and authenticating Unobtrusive Methods Unobtrusive methods involve researchers and research processes that are removed from the researched Unobtrusive methods are non-reactive and capitalize on existing data But researchers need to work through data not expressly generated for their proposes that may contain biases Unobtrusive Methods Unobtrusive methods include the exploration of official data and records corporate data personal records the media the arts social artefacts The Unobtrusive Process In order to gather data by unobtrusive means you need to know what you are looking for where you can find it whether it can be trusted what you can do with it Experimentation

Experimentation explores cause and effect relationships by manipulating independent variables in order to see if there is a corresponding effect on a dependent variable Experimentation Pure experimentation requires both a controlled environment and the use of a randomly assigned control group This can be difficult to achieve in human centred experiments conducted in the real-world Real-World Experiments There are many experiments that can only be carried out in the messy uncontrolled environments of the real-world, so the search for cause and effect will require tradeoffs between real-world contexts and a controlled environment

Once a research question has been agreed upon, a research design is formulated which includes an appropriate and effective method for collecting data. It has been rightly said by Sir Arthur Conan Doyle (Sherlock Holmes) that it is a capital mistake to theorize before one has data. In business research, data is collected from various sources and a variety of methods or techniques are used in its collection. It may be from a secondary source or primary source. In case of latter, it could be census or survey, laboratory experiment or field experiment, open or hidden observation. Each method has pros and cons. When population is large, it is neither necessary nor advisable to cover each and every member but a sample would suffice. It is quicker, cheaper and, if welldesigned, it can provide precise information about characteristics of the population. On the other hand, causal inference can only be drawn from experiments. So choice of the method depends upon time, money and objectives.

OBSERVATIONS
Observation is a primary method of collecting data by human, mechanical, electrical or electronics means with direct or indirect contact. As per Langley P, Observations involve looking and listening very carefully. We all watch other people sometimes but we do not usually watch them in order to discover particular information about their behavior. This is what observation in social science involves. Observation is the main source of information in the field research. The researcher goes into the field and observes the conditions in their natural state. There are many types of observation, direct or indirect, participant or non-participant, obtrusive or non-obtrusive, structured or non-structured. The observation is important and actual behavior of people is observed and not what people say they did or feel. For example, people value health but they would pick up food they know to fatty. It is useful when the subject cannot provide information or can only provide inaccurate information like people addicted to drugs. But at the same time, in observation the researcher does not get any insight into what people may be thinking.

OBTRUSIVE AND UNOBTRUSIVE

Obtrusive mean visible, thrusting out or evident. It is like class monitor, traffic warden or inspector. On the other hand, unobtrusive mean hidden, camouflaged or low-key. In such method, the researcher is not required to intrude. Also, it reduced bias. It enables to obtain data from a source other than people like instead of asking people which soft drink they like, an unobtrusive source

would to collect empty cans from garbage dumps and analyse their brands. There are other such examples: to ascertain popularity of journal, one can observe its wear and tear in a library. Also, entry counters in a super market provide very strong evidence that from which side the customers come in. Likewise, number of hits on a web site can be related to its usefulness or popularity. Hidden cameras can show consumer behavior in the store.

PARTICIPATIVE OR NON-PARTICIPATIVE
In participative observation, the observers becomes a participant in the program or culture or context being observed. It may require long time as the researcher needs to become accepted as a natural part of the culture. On the other hand, if one needs to observe child-mother interactions, one would resort to non-participative observations looking from a one-way mirror to note verbal or non-verbal cues being given by the mother and the child.

IN-DEPTH TECHNIQUES
In a survey, usually general questions are asked to know what customers or subjects do and think. But it one wants to know why they feel that way, one has to conduct an in-depth research. In survey, answers may depend on the mood of the respondent. As such, the survey shows how one feels at one particular time. But in in-depth research, long and probing interviews are taken to find out customers satisfaction and loyalty, usage, awareness and brand recognition etc. as discussed below:

FOCUS GROUP
A group is formed of 8 to 12 persons. They are selected keeping in view the targeted market. The group members are asked about their attitude towards a concept, product, service, packing or advertisement. Questions are asked in an interactive-group-setting where members are free to talk to each other. A moderator guides through the discussion. Through one-way mirror, the client or its representative observes the discussion, interpret facial expression and body language. There are some draw-backs. There is lesser control of the moderator or researcher and it lead to irrelevant discussions. Moreoveer, individual members consciously or unconsciously conform to what they perceive to be the consensus of the group, a situation called Group-think. The technology has give rise to Modern Group where group-members participate on-line and can share financial and operating data, pictures, voices and drawings etc.

PANELS
These are more or less like Focus Group. But Focus groups are formed for one-time discussion to decide about a particular issue. On the contrary, panels are of long-term nature for meeting frequently to resolve an issue. These can be current customers or potential customers; can be static or dynamic (members coming and going). Another difference is that penal are selected by the organizers through certain criteria like education, exposure and interest. Of course, there are exception like Microsoft Panel for Research and Evaluation of their software which are formed through open invitation.

IN-DEPTH INTERVIEWS

In-depth interviews, also called as one-to-one interviews, are expensive in term of time and money but are good for exploring several factors. Problems identified in an interview may be a symptom of a serious problem. The interviews may be conducted face to face or through telephone or these could be computer assisted interviews. Nowadays, television interviews have become more common. But interviews are fraught with bias from three sources, the interviewer, the interviewee and the interview setting. The interviewer may misinterpret the response or distort it while writing down. He may unintentionally encourage certain responses though gestures and facial impressions. The interviewee may not give his or her true opinion or avoid difficult questions. The setting may be good or bad creating comfort or discomfort. It may be open or in presence of some colleagues or senior or level of trust may be inadequate. In order to minimize bias, the interviewer should have knowledge, skill and confidence. Rapport and trust should be established in the interview.

PROJECTIVE METHODS
A psychological test in which a subject's responses to ambiguous or unstructured standard stimuli, such as a series of cartoons, abstract patterns, or incomplete sentences, are analyzed in order to determine underlying personality traits and feelings. This entails indirect question which enables the respondent to project beliefs and feelings onto a third party". The respondents are expected to interpret the situation through their own experience, attitude and personality and express hidden opinion and emotions. Of many techniques, word association, sentence completing and ink-blot tests are very common. In these techniques, both verbal and non-verbal (hesitation, time-lag and facial expression) are noted and interpreted. Such tests are useful for finding out consumer preference, buying attitude and behavior. Eventually, these are used for product development or finding out reason for failure of an apparently efficient product.

Presentation II: IN-DEPTH TECHNIQUES

EXPERIMENT FIELD AND LAB

Considerable data is generated or collected through experimentation in business research. A bank may conduct an experiment to know what attracts depositors: profit or security or liquidity. In a lab experiment, a group of 50 participants would be given chips representing money and would be shown three banks of which one was giving highest profitability, second was more "secure" and third provided liquidity through ATM and quick cheque processing system. The participants would be asked to deposit their chips with any of them. The total with each bank would be calculated and chips returned to participants. In the next round, the three banks would be rated equal expect one would be giving high returns and the participants would be asked to make their deposits. The experiment would go on and the data generated may point out a cause-and-effect relationship between any of the three motives and deposits. In case of field experiment, a bank may advertise that one of its branches is celebrating 100thanniversary and is offering 3% over and above its normal returns of 6%. If high return attracts more deposits, the clients would shift their saving towards that branch but if clients feel convenience is more important, there would not change in deposit levels.. One of the differences between lab and field experiment is use of real players or mock players. In other words use of actual clients or volunteers.

Survey

Surveys are based on short interviews either face-to-face or on telephone. These are based on simple questions. Also, there are computer-assisted interviews. Survey can be performed by sending a questionnaire and quantifying the response when questionnaires are returned.

Conclusion
Methods of collection data depends upon (i) nature of problem, and (iii) time and money available. Mostly, data is collected through use of secondary source. If no such data is available from liberaries or on the internet, one has to collect primary data for which a number of methods are available such as observations, indepth techniques, experiments and surveys.