STATISTICAL QUALITY

CONTROL

CHAPTER TWO:
Statistical Quality Control
tools
 Statistical Quality Control Tools
Quality control
 Quality control may generally be defined as a
system that maintains a desired level of quality,
through feedback on product/service
characteristics and implementation of remedial
actions, in case of a deviation of such
characteristics from a specified standard.
 This general area may be divided into three
main subareas: off-line quality control, statistical
process control, and acceptance sampling plans.
Quality improvement means the systematic
elimination of waste.
Off-Line Quality Control
 Off-line quality control procedures deal with
measures to select and choose controllable
product and process parameters in such a way
that the deviation between the product or
process output and the standard will be
minimized.
 The goal is to come up with a design within the
constraints of resources and environmental
parameters such that when production takes
place, the output meets the standard.
 Thus, to the extent possible, the product and
process parameters are set before production
begins.
Statistical process control
 Statistical Process Control is the application of
statistical techniques to control a process;
measuring and analyzing the variation in
processes so that performance remains within
specification limits.
 Process Improvement
 Is basically about the reduction of variability in
processes.
 It means work on the process to detect the
special causes of variation and remove them to
adjust the process.
 It results in waste reduction, quality
improvement, Customer Satisfaction.
Quality improvement methodologies
 A number of alternative and complementary
improvement methodologies have emerged,
the most widely publicized of which have
been proposed by American 'gurus':
 The Deming cycle;
 The Juran trilogy;
 The Crosby process.
 The Deming’s PDSA
cycle
Plan Stage:- (Identify problem
and develop plan for improvement)
 The current situation is investigated
and analyzed, data is collected to
understand fully the nature of the
problem being solved, and plans
made for improvement
 Predict the effect this change will
have and plan how the effects will be
measured.
Do Stage:- (Implement the change on a
small scale and measure the effects)
 Involves some form of trial or pilot
solution, for example with a small
part of a manufacturing or service
process or small group of customers.
 Study Stage:-
 In this stage the trial is
critically evaluated and
problems or other opportunities
are examined.
 Review the test, analyze the
results and identify what
you’ve learned.

Act Stage:-
 Ensures that the improvement
is implemented in a
standardized and continuous
Themanner
focus isbefore embarking
therefore continuous and closed-loop
upon the next plan.
improvement.
Problem Solving Cycle
Quality Trilogy Process
 Juran, like Deming, emphasized the need for
continuously working upon quality
improvements and encouraged the concept
of ‘breakthrough' whereby organizations
would achieve improvements leading to new
and unprecedented levels of performance.
 Stages of the Juran trilogy
I. Quality planning,
 The role of quality planning is to design a process
that will be able to meet established goals under
operating conditions.
 Quality planning involves:-
 Identifying customers, both internal and external
 Determining their needs
 Specifying the product features that satisfy those
needs at minimum cost.
 Designing the processes that can reliably
produce those features.
 Proving that the process can achieve its goals
under operating conditions
II. Quality control,
 The process of managing operations to meet quality
goals.
 Stresses the regular use of statistical control methods
to ensure that quality standards are met and to
identify variations from the standards.
 The process of Quality Control involves:
 Choosing control subjects
 Choosing units of measurement
 Establishing a measurement procedure
 Measuring
 Interpreting differences between measurement and goal
 Taking action to correct significant differences
III. Quality improvement,
 Quality improvement deals with the continuous
improvement of the product and the process.
 The steps in Quality Improvement:
 Prove the need for improvement
 Identify specific projects for improvement
 Organize to guide the projects
 Organize for diagnosis -- discovery of causes
 Diagnose the causes
 Provide remedies
 Prove that the remedies are effective under operating conditions
 Provide for control to maintain the gains.
 According to Juran, quality improvements should be
continuous as well as breakthrough.
The Juran trilogy
The Crosby’s process
 Crosby's approach to quality improvement
is based upon what he describes as the four
'absolutes' of quality, namely:
 Definition - understanding that quality means
conforming to the requirements.
 System - the approach to be adopted should
focus upon prevention rather than inspection.
 Performance standard - the organization should
strive for zero defects rather than adopt
acceptable quality (defect) levels.
 Measurement - the true measurement of quality
is the cost of non-conformances.
Quality Control Tools
The key process monitoring and
investigating tools include:
 Pareto charts,
 Cause and effect diagrams,
 Histograms,
 Check sheets,
 Flow charts,
 Scatter diagrams, and
 Control charts.
 Pareto chart
 A graphical tool for ranking causes from most
significant to least significant so that priorities for
process improvement can be established.
 It indicates which causes to tackle first by
showing the proportion that each cause
contributes to the total problem.
 The Pareto diagram is based on 80/20 principle,
which states that 80% of problems are created by
20% of causes.
Dr. Juran coined the phrases
“vital few and useful many”
 Steps to construct a Pareto diagram
Step 1: Determine the categories and the units for
comparison of the data, such as frequency, cost, or
time.
Step 2: Total the raw data in each category, then
determine the grand total by adding the totals of
each category.
Step 3: Re-order the categories from largest to
smallest.
Step 4: Determine the cumulative percent of each
category
Step 5: Draw and label the left-hand vertical axis with
the unit of comparison, such as frequency, cost or
time.
 Pareto Diagram
(cont’d)
Step 6: Draw and label the horizontal axis with the
categories. List from left to right in rank order.
Step 7: Draw and label the right-hand vertical axis from
0 to 100 percent.
Step 8: Beginning with the largest category, draw in
bars for each category representing the total for that
category.
Step 9: Draw a line graph beginning at the right-hand
corner of the first bar to represent the cumulative
percent for each category as measured on the right-
hand axis.
Step 10: Write any necessary items on the diagram.
Step 11: Analyze the chart. Usually the top 20% of the
categories will comprise roughly 80% of the cumulative
total.
 Pareto Diagram
(cont’d)
Example 6.2
The following table shows the different
types of defect and the total number of
items that are occurred on selected
products in an ideal company ABC. Use the
Pareto analysis to determine the vital few
cause, which results the majority of the
problem.
Table: Number of defects
observed
Type of Number of
Defect Defects
Crack 10
Scratch 42
Stain 6
Strain 104
Gap 4
Pinhole 20
Others 14
Total 200
 Table : Data Sheet for Pareto Diagram

Type of Number Cumulativ Percentag Cumulativ

Defects of Defects e Total e of e
overall Percentag
Total e
Strain 104 104 52 52
Scratch 42 146 21 73
Pinhole 20 166 10 83
Crack 10 176 5 88
Stain 6 182 3 91
Gap 4 186 2 93
Others 14 200 7 100
Total 200 - 100 -
 200 100
180 90

Number of defective units

Cumulative Percentage
160 80
140 70
120 60
100 50
80 40 A - Crack
60 30 B - Scratch
C - Stain
40 20 D - Strain
E - Gap
20 10 F - Pinhole

D B F A C E Others

Figure : Pareto Diagram by

Defective Items
 Cause and effect diagram
 Commonly called “fishbone diagram” because of
its appearance.
 Cause-and-effect diagrams are used to identify
and systematically list various causes that can
be attributed to a problem (or an effect).
What’s it used for?
· Identify the essential causes
of a problem
· Represent the relationships
between an “effect” and all its
possible “causes”
· Address the root causes of a If you
problem instead of just don’t get
treating the symptoms
the root,
The major purpose of this diagram the weed
is to act as a first step in problem
comes
solving by creating a list of possible
causes.
back!
 Causes are usually grouped in to major
categories to identify these sources of
variation.
 The categories typically include:
 People (Man): any one involved in the process.
 Methods: how the process is performed .
 Machines: any equipment, computers, tools, etc. required
 Materials: raw materials, parts, etc. used to produce the
final product.
 Measurements: data generated from the process that are
used to evaluate its quality.
 Environment: the conditions, such as location, time,
temperature, and culture in which the process operates.
 Creating Fishbone Diagrams
As a group:
1. Establish problem (effect)
-state in clear terms
-agreed upon by entire group
2. Problem becomes the “head” of the fish
-draw line to head (“backbone”)
3. Decide major causes of the problem
- by brainstorming
4. Connect major causes to backbone of the fish with slanting
arrows
5. Brainstorm secondary causes for each of the major causes
6. Connect these secondary causes to their respective
major causes
7. Repeat steps 5 & 6 for sub-causes dividing with increased
specificity
- usually four or five levels
8. Analyze and evaluate causes and sub-causes
- may require the use of statistical, analytical, and
graphical tools
 Cause and Effect Diagram
Step 1 & 2:

Quality
Quality
Problem
Problem

“backbone”
“head”
 Cause and Effect Diagram
Step 3 & 4:

Measurement
Measurement Human
Human Machines
Machines

Quality
Quality
Problem
Problem

Environment
Environment Materials
Materials Process
Process
 Cause and Effect Diagram
Step 5, 6, & 7:

Measurement
Measurement Human
Human Machines
Machines
Faulty
testing equipment Poor supervision Out of adjustment

Incorrect specificationsLack of concentration Tooling problems

Improper methods Inadequate training Old / worn

Quality
Quality
Inaccurate Problem
Problem
temperature
Defective from vendor Poor process design
control
Ineffective quality
Not to specifications
management
Dust and Dirt Material- Deficiencies
handling problems in product design

Environment
Environment Materials
Materials Process
Process
 Step 8: use tools to analyze and evaluate causes
 Pareto diagrams, charts, and graphs
 Statistical analysis for causes in processes

 Step 9: decide and take action

 Use fishbone diagram, analysis and evaluations to find
causes that can be fixed
 Take action to eliminate and fix problem causes
 Histogram
 The histogram is a bar chart showing a
distribution of variables (variation in a
process).
 It is a graphic summary of the distribution of
a set of data.
 A histogram is a snapshot of the variation of a
product or the results of a process. It often
forms the bell-shaped curve which is
characteristic of a normal process.
 The histogram clearly portrays information on
location, spread, and shape regarding the
functioning of the physical process.
The rationale behind the
histogram:
 Almost any set of data will show variation.
 This variation will exhibit some pattern.
 The pattern can give us clues as to what is
happening in the process from which the
data were obtained.
 It is difficult to see this pattern in a table of
data.
 The histogram presents the variation in a
way that helps us understand what may be
happening in the process.
A useful tool because the analyst can
quickly visualize the features of the
data, such as:
 Shape of the distribution
 Any central tendency in the distribution
or mean
 Amount of variation or spread
 Range of measurements which aids in
the process capability
Steps in constructing Histogram

A) HOW TO MAKE FREQUENCY TABLES

Step 1: Calculate the range (R)
R = (the largest observed value)-(the smallest
observed value)
Step 2: Determine the class interval
 The class interval is determined so that the
range, which includes the maximum and the
minimum of values, is divided in to intervals of
equal breadth.
 To obtain the interval breadth, divide R by
odd intervals of factor 1, 3, or 5 (or 10, 30,
50; 0.1, 0.3, 0.5 etc.) so as to obtain from 5
to 20 class intervals (cells) of equal
breadth.
 As a guideline, use 5 to 9 cells when the
no. of observations is less than 100; use 8
to 17 cells when the no. of observations is
between 100 and 500; and use 15 to 20
cells when the no. of observations is
greater than
i 
500.R
1  3.322 log n
 Or, use the simplest technique, Sturgi’s
Step 3: Prepare the frequency table
Prepare a table in which the class, midpoint,
frequency marks frequency, etc., can be
recorded.

Step 4: Determine the class boundaries

Determine the boundaries of the intervals so
that they include the smallest and the largest
of values, and write these down on the
frequency table.
 Step 5: Calculate the mid point of the class
Using the following equation, calculate the
mid-point of class, and write this down on the
frequency table.Sum of the upper & lower boundaries of the 1st class
Mid  point of the 1st class 
2

Similarly, Sum of the upper & lower boundaries of the 2nd class
Mid  point of the 2nd class 
2
  The mid points of the second class, the third
class, and so on, may also be determined as
follows:
Mid  point of the second class Mid  point of the first class  class interval

Mid  point of the third class Mid  point of the second class  class interval

and so on.

Step 6: Obtain the frequencies

Read the observed values one by one and
record the frequencies falling in each class
using tally marks, in group of five.
Example 6.1
Mesfin Industrial Engineering Share
Company wants to investigate the
distribution of the diameters of shafts
produced in a grinding process, the diameter
of 90 shafts are measured as shown in the
following table. Draw a histogram using
these data.
 Table : Sample and Result of Measurement
Sample Results of measurements
Number

1-10 2.510 2.517 2.522 2.522 2.510 2.511 2.519 2.532 2.543 2.525

11-20 2.527 2.536 2.506 2.541 2.512 2.515 2.521 2.536 2.529 2.524

21-30 2.529 2.523 2.523 2.523 2.519 2.528 2.543 2.538 2.518 2.534

31-40 2.520 2.514 2.512 2.534 2.526 2.530 2.532 2.526 2.523 2.520
41-50 2.535 2.523 2.526 2.525 2.523 2.522 2.502 2.530 2.522 2.514

51-60 2.533 2.510 2.542 2.524 2.530 2.521 2.522 2.53 2.540 2.528

61-70 2.525 2.515 2.520 2.519 2.526 2.527 2.522 2.542 2.540 2.528

71-80 2.531 2.545 2.524 2.522 2.520 2.519 2.519 2.529 2.522 2.513

81-90 2.518 2.527 2.511 2.519 2.531 2.527 2.529 2.528 2.519 2.521
Solution:
Step 1: Calculate R
R is obtained from the largest and the
smallest of observed values. Therefore; from
the above table:
The largest value is 2.545
The smallest value is 2.502
Thus, R = 2.545 - 2.502 = 0.043
Step 2: Determine the class interval
0.043/0.001 = 43, and we can take this 43 as it is
0.043/0.003 = 14.33, and we can make this 14 by
rounding down to the nearest integer
0.043/0.005 = 8.6, and we can make this 9 by
rounding up to the nearest integer
0.043/0.007 = 6.14, and we can make this 6 by
rounding down to the nearest integer.
We know that n = 90; hence, the class interval is
determined as 0.005 or 0.007, since these give a
number of intervals between 5 and 9. But, we
generally take the narrower interval for less than
100 observations and the wider for more than 100
observations. Hence, we take 0.005.
Step 3: Prepare the frequency table
Prepare a table as shown in the Table below

Step 4: Determine the class boundaries

The boundaries of the first class should be
determined as 2.5005 and 2.5055 so that the
class includes the smallest value 2.502; the
boundaries of the second class should be
determined as 2.5055-2.515, and so on.
Record these on frequency table.
Step 5: Calculate the mid-point of class

2.5005  2.5055
Mid point of the first 
class  2.503
2
2.5055  2.5105
  2.508
Mid point of the second class 2
and so on.

Step 6: obtain the frequencies

Record the frequencies.
 Table: Frequency Table

Class Mid- Frequency mark Frequency

point of (tally) f
class x
1 2.5005-2.5055 2.503 / 1
2 2.5055-2.5105 2.508 //// 4
3 2.5105-2.5155 2.513 ///// //// 9
4 2.5155-2.5205 2.518 ///// ///// //// 14
5 2.5205-2.5255 2.523 ///// ///// ///// ///// // 22
6 2.5255-2.5305 2.528 ///// ///// ///// //// 19
7 2.5305-2.5355 2.533 ///// ///// 10
8 2.5355-2.5405 2.538 ///// 5
9 2.5405-2.5455 2.543 ///// / 6
Total 90
 B) HOW TO MAKE A HISTOGRAM
Step 1:
On a sheet of squared paper, mark the
horizontal axis with a scale. The scale should
not be no the base of class interval but it is
better to be on the base of measurement of
data, (e.g. 10 grams correspond to 10 mm).
Step 2:
Make the left-hand vertical axis with a frequency
scale, and, if necessary, draw the right-hand axis
and mark it with a relative frequency scales.
Step 3:
Make the horizontal scale with the class boundary
values.
Step 4:
Using the class interval as a base line, draw a
rectangle whose height corresponds with the
frequency in that class
Step 5:
Draw a line on the histogram to represent the
mean, and also draw a line representing the
specification limit, if any.
Step 6:
In a blank area of the histogram (Figure below),
note the history of the data.
Figure: Histogram for the
above Example
Figure: Histogram for the
above Example
 Flow Chart
A flow chart is a pictorial representation
showing all of the steps of a process.
Check Sheet
A check sheet is a paper form on which items
to be checked have been printed already so
that data can be collected easily and
concisely.
Its main purposes are:
 To make data-gathering easy
 To arrange data automatically so that
they can be used easily later on.
 Check sheets come in several types,
depending on the objective for collecting the
data.
 Some of th e more common ones are as
follows:
 A. Recording check sheet
 Characteristic: Customer Complaints on Refrigerators
 Period of data collection: October 2017
 Source of data: logbook
B. Location check sheet
C. Checklist check sheet
 Scatter Diagram
The scatter diagram is a technique used to
study the relation of two corresponding
variables.

The two variables we will deal with are:

1. A quality characteristic and a factor
affecting it,
2. Two related quality characteristics, or
3. Two factors relating to a single quality
characteristic.
 Interpretation of scatter diagrams
Degree of correlation
 The correlation coefficient computed from the
sample data measures the strength and direction of a
relationship between two variables.
 The symbol for the sample correlation coefficient is r;
Karl Pearson’s coefficient of correlation.
N  XY    X Y
r 
N  X 2
  X   N  Y    Y  
2 2 2

Where, N is the number of data

pairs
 r value close to - 1 indicates a strong negative correlation.
 r value close to + 1 indicates a strong positive correlation.
 r value close to 0 indicates that the correlation between
the two characteristic is weak.
 Interpretation of scatter diagrams
The scatter diagram may show that two
variables have positive correlation, negative
correlation, or no correlation at all.

58
Fig - Common Scatter Diagrams
 Scatter
(cont’d)
Example 6.4
A manufacturer of plastic tanks who made them
using the blow molding method encountered
problems with defective tanks that had thin tank
walls. It was suspected that the variation in air
pressure, which varied from day to day, was the
cause of the non- conforming thin walls.
Construct a scatter diagram to decide the
relation ship between the two variables? Also
determine the Karl Pearson’s coefficient of
correlation (r) and comment on the value
obtained?
 Scatter (cont’d)

Table below shows data on blowing air-pressure

and percent defective. Let us draw a scatter
diagram using this data, according to the steps
given above.

Step 1:
As seen in Table below, there are 30 pairs of
data.
 Table : Variations in Air Pressure

No. Air Percent No Air Percent

Pressure Defective[ . pressure Defective
[kgf/cm2 %] [kgf/cm2] [%]
1 8.6 0.889 9 9.2 0.895
2 8.9 0.884 10 8.7 0.896
3 8.8 0.874 11 8.4 0.894
4 8.8 0.891 12 8.2 0.864
5 8.4 0.874 13 9.2 0.922
6 8.7 0.886 14 8.7 0.909
7 9.2 0.911 15 9.4 0.905
8 8.6 0.912 16 8.7 0.892
No. Air Percent No. Air Percent
Pressure Defective[ pressure Defective
[kgf/cm2] %] [kgf/cm2] [%]
17 8.5 0.877 24 8.9 0.908

18 9.2 0.885 25 8.3 0.881

19 8.5 0.866 26 8.7 0.882

20 8.3 0.896 27 8.9 0.904

21 8.7 0.896 28 8.7 0.912

22 9.3 0.928 29 9.1 0.925

23 8.9 0.886 30 8.7 0.872

Step 2:
Blowing air pressure is indicated by x (horizontal axis),
and percent defective by y (vertical axis). Then,
the maximum value of x: xmax = 9.4 (kgf/cm2) ,
the minimum value of x: xmin. = 8.2 (kgf/cm2),
the maximum value of y: ymax = 0.928 (%),
the minimum value of y: ymin. = 0.864 (%).
We mark off the horizontal axis in 0.5 (kgf/cm2)
intervals, from 8.0 to 9.5 (kgf/cm2), and the vertical
axis in 0.01 (%) intervals, from 0.85 to 0.93 (%).
 Scatter
(cont’d)

Step 3: Plot the data. (See Figure below)

Step 4: Enter the time interval of the sample

obtained (Oct. 1 -Nov. 9), number of samples
(n = 30), horizontal axis (blowing air-pressure
[kgf/cm2]), vertical axis (percent defective
[%]), and title of diagram (Scatter diagram of
blowing air-pressure and percent defective).
 Scatter
(cont’d)

Figure : Scatter Diagram of Blowing Air Pressure and Percent Defective

• Determine the Karl Pearson’s coefficient of correlation (r) and
comment on the relation ship between the two variables?