7 QC Tools

Check-sheet

Cause and Effect

Pareto

Histogram

Control Chart

Scatter Plot

Stratification

Check-sheet

For problem solving data is to be captured.

Check-sheet is a tool to capture data as per

check list

Check list is a tool to capture the parameter

4. Additional causes can be branched off the tertiary causes.

Pareto Chart
Find the few important reasons !
the vital few and the trivial many
80 20 rule
Class data

Collect

data in a table

Calculate
Plot

the cumulative values

pareto diagram

Example
Type of Defect
No. of defective pc. %
Sheet Hard
56
44.1
Draw & Die Setting
49
38.2
Sheet Line
15
11.8
Operator Defect
7
5.9
Total
127

Cum. %
44.1
82.3
94.1
100

PARETO CHART FOR PRESS SHOP (DEC-04)

120

46.1

45

38.5

40

100

84.6

35
% AGE

94.9

100
80

30
25
20

60

46.1

15

40

10.3

10

5.1

20

5
0

0
Sheet Hard

Draw & Die

Setting

Sheet Line

DEFECTS

Operator Defect

CUMULATIVE %AGE

50

What is a Histogram?

Histogram is a visual tool for presenting variable

data. It organises data to describe the process
performance.

Additionally histogram shows the amount and

pattern of the variation from the process.

Histogram offers a snapshot in time of the process

performance.

Definition of Histogram

A histogram is a graphical summary

of variation in a set of data.

The pictorial nature of the histogram

enables us to see patterns that are difficult
to see in a table of numbers.

Key Concept of Histogram

Data always have variation

Variation have pattern

Patterns can be seen easily when

summarized pictorially

While studying histogram

look for its

Location of mean of the process

Spread of the process

Shape of the process

Calculations for Histogram

1

Xs

Xl

51

50

50

53

48

48

53

55

49

49

50

49

49

55

49

48

51

53

49

48

53

50

50

50

51

53

50

53

49

50

51

47

51

47

51

50

51

52

50

50

50

52

51

51

52

48

50

48

52

54

49

50

49

50

49

54

51

50

50

53

50

50

53

10

49

51

53

49

51

49

53

Calculations for Histogram

Smallest Value, S= 47

Largest Value, L = 55

Range= L-S= R= 8

No. of cells= 1+3.22log10(50)= 7

Calculated cell width (CW)= R/no. of cell=1.14

Rounded off Cell width= 1 (multiple of 1,2,5 of

least count

Calculations for Histogram

Starting value, A= 47

LCB(1)= A-cw/2= 47-1/2= 46.5

UCB (1)= LCB(1)+CW= 46.5+1= 47.5

Plotting Histogram
Lower

Upper

Mid
value

Tally
Mark

46.5

47.5

47

47.5

48.5

48

III

48.5

49.5

49

10

49.5

50.5

50

16

50.5

51.5

51

11

51.5

52.5

52

52.5

53.5

53

53.5

54.5

54

54.5

55.5

55

Tally
Mark

Freq

18
16
14
12
10
8
6
4
2
0

Histogram

Skewness
Is the histogram symmetrical? If so, Skewness is zero. If
the left hand tail is longer, skewness will be negative. If
the right hand tail is longer, skewness will be positive.
Where skewness exists, process capability indices are
suspect.
For process improvement, a good rule of thumb
Kurtosis
is to look at the long tail of your distribution; that is
Kurtosis
is a measure
of the pointiness
of a distribution.
usually where
quality problems
lie
The standard normal curve has a kurtosis of zero. The
Matter horn, has negative kurtosis, while a flatter curve
would have positive kurtosis. Positive kurtosis is usually
more of a problem for quality control, since, with "big"
tails, the process may well be wider than the spec limits.

Distributions you may encounter

The standard normal distribution, with its zero skewness and zero
kurtosis

A skewed distribution, with one tail longer than the other.

A double-peaked curve often means that the data actually reflects

two distinct processes with different centers. You will need to
distinguish between the two processes to get a clear view of what
is really happening in either individual process

A truncated curve, with the peak at or near the edge while trailing
gently off to the other side, often means that part of the distribution has
been removed through screening, 100% inspection, or review. These
efforts are usually costly and make good candidates for improvement
efforts.

A plateau-like curve often means that the process is ill-defined to those doing the
work, which leaves everyone on their own. Since everyone handles the process
there are many different measurements with none standing out. The
differently,
solution here is to clearly define an efficient process.

Outliers in a histogram bars that are removed from the others by at least the
width of one bar sometimes indicate that perhaps a separate process is included,
but one that doesn't happen all the time. It may also indicate that special causes of
variation are present in the process and should be investigated, though if the
process is in control before the histogram is made as it should be, this latter option
is unlikely

Why Control Chart ?

To find

Is there any change in location of process

average ?

Is there any change in the spread of the process ?

Is there any change in shape?

Control Charts
Variables

Attributes

R Chart
X s Chart

p Chart
np Chart
C Chart
u Chart

Variable Control Charts Attribute Control Charts

o

o
o
o

defect prevention and

process improvement
more expensive to
construct and maintain
can tell reason for
process behavior
smaller n (1-10) needed

o
o

o
o
o

defect detection
a screening device to
initiate variables control
charting
cheaper to construct and
maintain
cannot tell cause of
defect
need large n (>100)

Attribute Control Charts

p Chart

np Chart

Measures number of defective pieces

Charts the numbers of defective pieces in fixed size samples

C Chart

Measures % defective
Charts number of defects in varying sized samples

Measures number of defects

Looks at a single product or piece

u Chart

Measures number of defects per unit area, time, length, etc.

Charts number of defects in a product of varying size

Selecting a Control Chart

Start
Measurable
Variables Data

n=1

c or u
Chart

S.D,
if n>10

X-bar s Chart

Constant
n

No
u
Chart

Yes
np or p
Chart

Defectives

n>1

X-bar R Chart

Defectives
or
Defects?

Defects
Yes

Range or
S.D

Attributes Data

DATA TYPE

Sample
Size, n

X-bar
MR Chart

Range,
if n<10

Countable

Constant
n

No
p
Chart

Control Limits
Upper Control Limit
Target
3 x sd of means
Lower Control Limit
1

Sample Number

Control Chart Technique - 1

Select a quality characteristics
Weight
Length
Viscosity
Tensile

Strength

Capacitance

Control Chart Technique - 2

Choose sub group size
Sensitivity increases with the sub group size
Cost of sampling increases with size
In case of destructive testing - 2 or 3
Normally sub group size can be 4 or 5
Choose Interval of data collection (frequency)
of sample size
Collect the data

If mean of the process shifts by 1

times sd ( Impact of sample size)
Sample Size

Chances of Detecting a
shift (%)

10

10

55

15

82

20

95

Typical Data Table

Operation
SN

Date

Other Details
Measurement

Time

Mean Range

X1

X2

X3

X4

12/12 10.25

35

40

32

33

35.0

12/12 13.45

46

42

40

38

41.5

12/12 15.34

34

40

34

36

36.0

15/12 10.30

38

34

44

40

39

10

..

Determine trial control limits

Constants for Trial Control Limits

group Size

A2

D4

D3

1.880 3.267

1.023 2.527

0.729 2.282

0.577 2.115

0.483 2.004

0.419 1.924 0.076

Formula for Control Limits

For mean control chart

Upper Control Limit, UCLx = T + A2 x R
Lower Control Limit, LCLx = T - A2 x R
For range control chart
Upper Control Limit, UCLr = D4 x R
Lower Control Limit, LCLr = D3 x R

Sample Calculation

In our case
Target, T = 50
Mean range, R = 9
Sub Group size, n = 4

Values obtained from the table of constants

A2 = 0.729
D4 = 2.282
D3 = 0

Sample Calculations

UCLX = 50 + 0.729 X 9 = 56.56

LCLX = 50 - 1.079 X 9 = 43.44
UCLr = 2.282 X 9 = 20.54
LCLr = 0 X 9 = 0

Outliers

Outliers are those observations which do

not belong to normal population.

If Outliers are included in the calculation,

then the information is distorted. Not more
than 20% subgroups are omitted

Outliers

Scan column for means

If any mean is more than UCLx or mean is
less than LCLx then drop that sub-group

Checking for range outliers

Scan column for ranges, if any range is more
than UCLr then drop that sub-group

If any sub-group(s) is dropped then

recalculate the trial control limits using
remaining sub-group(s)

Continue this exercise till there is no

further droppings ( max 20%)

Mean

60
55
50
45
40

Range

Control Chart

20
10
0

UCLx
T=50
LCLx
UCLr

Sample Number

Mean

Control Chart with Action Limit

60
50
40
30
40
30
20
10
0

UCLx
UWLx

2/3 A2 x R

T=50

2/3 A2 x R

Sample Number

LWLx
LCLx

One Point Falling Out side Control Limit

Sample Mean

UCLx

LCLx

Sample Number

Sample Mean

Two out of 3 points falling between Control Limit

and Warning Limits
UCLx

LCLx

Sample Number

Sample Mean

7 Consecutive points falling on one side of

the center line ( A run of seven )

UCLx

LCLx

Sample Number

Over Adjustment

Over adjustment is fiddling with the controls of

stable process.

It is often a well intention move of the process

owner, but bad attempt to improve the process.

It actually adds a further source of variation to the

process, and hence will increase total variation .

Problem of Over Adjustment

Resultant distribution
with flat top

38 40 42 44 46 48 50 52 54 56 58

Scatter Plot

Scatter diagram Exhibits

Relationship
A scatter diagram shows the
relationship between independent
variable (cause) and dependent
variable (effect).

The independent variable is plotted

on x-axis and dependent variable on yaxis.

Characteristics of Independent
Variable

It should be measurable on a continuous scale.

It should have a logical relationship with the

dependent variable.

Changes in level of independent variable

should cause changes in level of dependent
variable.

Typical Relationship We Normally Like to

Study

Independent Variable

Dependent Variable

Moisture contents

Elongation of thread

Wax purity

Hardness of lipstick

Roller Pressure

Paper thickness

Charge weight

Range of bullet

Number of users

Response time

Length of bar

Typical Relationship

X
Pull Speed

Typical
Relationship

Potency

Shelf Life

Table - Humidity Vs Voltage

Humidity %
10
20
30
40
50
60
70
80
90
100

V1
40
46
45
49
51
54
54
57
59
60

V2
43
43
43
45
47
51
52
55
57
58

Voltage
V3
41
46
43
48
50
51
51
54
56
57

V4
42
46
44
49
51
52
55
58
59
58

V5
40
44
43
46
49
53
53
58
57
58

60

Scatter Plot

Voltage

55
50
45
40
35
10 20 30 40 50 60 70 80 90 100
Humidity

Stratification
Stratification is simply the creation of a set of pareto
charts for the same data, using different possible
causative factors

Following figures plot defects against three possible sets of

potential causes. The figure shows that there is no significant
difference in defects between production lines or shifts. But
product type three has significantly more defects than do the
others. Finding the reason for this difference in number of
defects could be worthwhile

A problem is solved according to the

following SEVEN STEPS of Q.I.Story.
SEVEN STEPS OF Q.I.STORY :
1.Reason For Improvement
2.Current Situation
3.Analysis
4.Counter Measure
5.Result
6.Standardization
7.Future Plan