UNIT 1

OVERVIEW OF STATISTICAL
QUALITY CONTROL

Structure
1.1 Introduction Chance Causes

Expected Learning Outcomes Assignable Causes

1.2 Quality 1.5 Control Charts

Defining Quality 1.6 3σ Limits
Dimensions of Quality 1.7 Control Chart Patterns
Quality Control Natural Patterns of Variation

Historical Review of Quality Control Unnatural Patterns of Variation

1.3 Statistical Quality Control 1.8 Advantages and Limitations

of SQC
Elements of SQC
1.9 Summary
Techniques of SQC
1.10 Terminal Questions
Statistical Process Control
1.11 Solutions/Answers
Product Control

1.4 Causes of Variation

1.1 INTRODUCTION
In our day-to-day life, we use different products and services from the time we
get up until we get back to bed. For example, we use different types of
products such as toothpaste, soap, detergent, clothes, food, gas stove,
vehicles, phones, computers, TV, electric bulb, etc. and we also use different
types of services such as water supply, electricity, transportation, health care,
education, etc.
All of us wish that these products and services would fulfil certain
specifications when we use them. If the product/service fulfils the
specifications needed for its proper use, we say that it is of good quality. If it
does not fulfil the specifications, it is said to be of poor quality.
In the face of ever-growing market competition, the main objective of the
manufacturer or producer is to achieve quality assurance in manufacturing
9
Dr. Prabhat Kumar Sangal, School of Sciences, IGNOU, New Delhi
Block 1 Process Control
and service organisations so that his/her product/service can meet the existing
competition in the market. In order to achieve this objective, different statistical
tools have been developed, which are useful for controlling the quality of
products vis-a-vis certain specifications or standards. The technique of
controlling product quality against the specifications using statistical tools is
known as Statistical Quality Control (SQC).
In this unit, you will learn about the concepts and various aspects of SQC. In
Sec. 1.2, we define the term quality and discuss dimensions of quality, quality
control and historical review of quality control. In Sec. 1.3, we describe various
aspects of SQC, e.g., the elements and techniques of SQC, statistical process
control and product control. We also discuss the causes of variation, which
may be due to chance or could be assigned to some factors in the production
process in Sec. 1.4. In Sec. 1.5, we introduce the control chart - a tool used in
statistical quality control to indicate whether a process is under control or
out-of-control. We explain the concept of 3σ limits, different patterns of the
control chart and advantages and limitations of SQC in Secs. 1.5 to 1.8. In the
next unit, you will study the control charts for variables.
Expected Learning Outcomes
After studying this unit, you should be able to:
 explain the concepts of quality, quality control and statistical quality
control (SQC);
 describe the need for statistical quality control;
 distinguish between chance causes and assignable causes in the
production process;
 describe the techniques of SQC;
 define process control and product control;
 explain the concept of control chart, the principle underlying 3σ limits
and various control chart patterns; and
 discuss advantages and limitations of SQC.

1.2 QUALITY
In the introduction of the unit, we have just given you the flavour of the notion
of quality. We now define the term quality. The dictionary meaning of quality
is Degree of Excellence. It means that both products and services should be
excellent. This definition of the term quality is rather subjective as it depends
on the perception of excellence and varies from situation to situation or from
person to person. If 20 different people are asked to define quality, most
probably there will be 20 different answers. However, the best will be widely
used. If we ask what quality they want, then most of the people say that they
want the best. This indicates that the general public opinion about the term
quality is the best. However, if we observe the buying habits of people, we
find that most people buy an item at a discount price in a sale or a mid-priced
item rather than the most expensive one. This indicates that while buying
something, people want the best that they can afford. Further, apart from
price, there are many other things to be considered for quality such as size,
10 performance, warranty, appearance, etc.
Unit 1 Overview of Statistical Quality Control
The concept of quality as goodness or the best or luxury is not adequate for
professionals working in quality control because the best depends on an
individual’s perception. For example, a person may think that a car is a good
vehicle for Delhi roads, but a comparatively poor vehicle for the forests of
Himachal Pradesh, Madhya Pradesh, etc. because the best vehicle for these
places would be one with a four-wheel drive. Similarly, a good walking shoe is
good for walking but unsuitable for running. Therefore, the intended use of
the product or service needs to be included in the concept of quality.
Therefore, we need to define the term quality.
1.2.1 Defining Quality
There are several ways of defining quality:
1. Initially, quality was defined as conforming to specifications. This
means that any product should be manufactured according to the given
specifications. If a product fulfils its specifications, it is considered to be a
quality product.
There are some problems with this definition. In some cases, it is found
that even though a product conforms to all specifications, its utility is not
up to the mark. For example, suppose a person wishes to purchase a
touchscreen mobile phone having clear sound. The shopkeeper shows
him/her such a mobile set. But the person may feel that the size of the
mobile phone is big. So he/she may not purchase it because it is not fit for
his/her use. Therefore, from the point of view of customers, such products
are not useful. Thus, there is a need to redefine the term quality.
2. The definition of quality was modified to include fitness for use. This
means that the manufactured product should be such that it conforms to
its specifications and is fit for use. However, it was found that in some
cases, a product conformed to all specifications and was fit for use but
could not be sold as it did not appeal to the customer. Therefore, to sell
any product it is necessary to incorporate the customer’s viewpoint.
3. According to the third definition, quality is customer satisfaction. This
means that a product, which brings satisfaction to the customer could be
termed as a quality product. Then it was found that every customer would
have plenty of demands with respect to each product that he/she wished
to buy. However, customers seldom express all of their expectations. For
example, suppose a person goes to a restaurant and orders a pizza. If the
pizza given to him is not hot, that person would definitely be unhappy. But
it is also true that the customer would never mention that he/she wanted a
hot pizza! There are many situations where customers are unhappy, even
when all their expressed or stated needs are fulfilled. Hence, while
defining quality, there is a need to consider for considering the
unexpressed or unstated needs of customers.
4. The fourth definition given for quality was delighting the customer.
Delight is one step ahead of satisfaction. When a product fulfils both the
expressed and unexpressed needs of the customer, he/she is delighted.
However, this definition was also improved upon.
5. The fifth definition given for quality was enchanting the customers. 11
 Block 1 Process Control
According to this definition, the manufacturer plays a dual role: firstly,
he/she should know the needs of the customers. Secondly, he/she should
make customers aware of this fact and also make them feel that these are
the products they want. For example, an electric bulb manufacturer needs
to educate customers about LEDs and make them want to buy LEDs as
these are low on the consumption of electricity even though their cost
slightly more. Nowadays, every manufacturer is expected to follow this
definition.
From the above discussion, we may conclude that in the manufacturing/
services sectors, the following aspects have to be incorporated in the definition
of quality:
• conforming to specifications,
• fitness for use,
• customer satisfaction,
• delighting the customer, and
• enchanting the customers.
Having explained the concept of quality in industry, and defined it we now
describe various dimensions of quality.
1.2.2 Dimensions of Quality
David A. Garvin is the C. In the previous section, we have explained the concept of quality and defined
Roland Christensen it. In 1988, David A. Garvin summarised eight basic elements of quality, which
Professor of Business are known as the dimensions of quality. We describe them briefly.
Administration at the
(i) Performance
Harvard Business School,
Massachussetes USA. The first dimension of quality is performance. It refers to the primary
operating characteristics of a product. Consumers judge the quality of
any product based on its performance after comparing it with the
competitor’s products or the prevailing market standard. For example, a
mobile phone can be judged by its clarity of sound, weight, size,
functions, etc. Similarly, a motorcycle can be judged by its pick-up, fuel
efficiency, etc.

(ii) Features

Features constitute the second dimension of quality. These refer to

additional characteristics available in products along with the primary
operating characteristics. For example, complimentary drinks and
snacks in a flight or hotel, Bluetooth and FM in mobile, etc. A customer's
choice is also influenced by this dimension of quality.

(iii) Reliability

Reliability refers to the probability of a product’s failure within a

specified time period. If a product fails frequently, we say that it is an
unreliable product. For example, if a TV of a particular company requires
frequent repair, we say that it is unreliable. There are many products
such as laptops, TVs, automobiles, etc. in which customers’ view about
12
quality is influenced by the reliability of the product.
Unit 1 Overview of Statistical Quality Control
(iv) Conformance (Agreement)

Conformance means meeting specifications. Customers obviously want

that the product should meet its specifications. For example, when we
purchase a motorcycle, we check whether the sitting space, weight, size,
pick-up, fuel efficiency, etc. conform to the specifications mentioned by
the company. This dimension also influences customer's choices.

(v) Durability

Durability refers to the measure of product life. It can also be

understood as the operational life of the product, i.e., how long the
product can be used without replacement. The life of an electric bulb is
an example of durability: when its filament burns up, the bulb needs to
be replaced because at this stage, the repair is not possible. Durability
also changes customer's views.

(vi) Serviceability

Serviceability is the sixth dimension of the quality. Consumers are

concerned not only about a product breakdown but also about the time
taken before the product is serviced and restored. Serviceability is
concerned with how readily a product can be serviced back into
operational mode. For example, suppose a customer wishes to buy a
washing machine. While deciding on the company, he/she may consider
how long the company takes to service or repair it in case it breaks
down.

(vii) Aesthetics

The seventh dimension of quality is aesthetics, which means how a

product looks, sounds, feels, etc. The aesthetic value of a product is
purely subjective. For example, some people may find a particular car
aesthetically appealing while others may not.

(viii) Reputation

Reputation is related to the past performance of the company. In many

cases, customers also check out the quality of products made by the
company in the past. For instance, if a company launches a new car, it is
usually assumed by the customer that the new model will be successful
because the past performance of other cars of that company was good.

In order to design and manufacture products of high quality, it is necessary to

incorporate all eight dimensions of quality.

You may like to pause here and check your understanding of the definition of
quality and its dimensions by answering the following Self Assessment
Question.

SAQ 1
Choose the correct option from the following:
(i) Quality means
(a) fitness for use 13
Block 1 Process Control
(b) degree of excellence

(c) conformance to requirements or specifications

(d) all of the above

(ii) Primary operating characteristic of a product is known as

(a) aesthetics

(b) reliability

(c) performance

(d) features

So far you have studied the definition of quality and its dimensions. We now
introduce the concept of quality control.

1.2.3 Quality Control

We first understand what we mean by control.

The process/procedure/method that is applied to meet the specifications or

standard is known as control. Process control works on feedback and
comprises the following steps:

1. Choose the parameters to control, i.e., we first choose the characteristic

that we intend to control such as length, height, weight, defects, etc.

2. Choose the unit of measurement, e.g., centimetre (cm), millimetre

(mm), gram (g), etc.

3. Set the standard or specification for the parameter to control, e.g., 5

cm, 10 g, etc.

4. Select a sensing device, which can measure the parameter to control in

terms of the unit of measure, e.g., scale, a weighing balance, etc.

5. Measure the actual performance.

6. Compare the actual performance with the standard.

7. Take necessary action when there is a difference between actual

performance and the standard.

Thus, quality control can be defined as:

“The process by which we measure the quality characteristics of the product,

compare them with the specifications or standard and take suitable actions
whenever there is a difference between actual quality and the specifications or
standard”.

Nowadays, quality is controlled by using statistical tools. This technique is

known as Statistical Quality Control, which we describe in the next section.
However, before you study further, you may like to know how the concept of
14 quality control evolved historically.
Unit 1 Overview of Statistical Quality Control
1.2.4 Historical Review of Quality Control
The notion of quality control is not new. We may say that it began when
human beings lived in forests and ate raw plants and animals. In those times,
they used natural materials and may have faced problems such as which
plants were suitable for eating and which ones were poisonous to health. With
the evolution of early technology, human beings started processing natural
materials. They made products and examined whether these worked properly
for their purpose or not. For example, they produced rough-hewn stone tools.
They may have checked the point of the tool to see whether it was sharped
enough for its purpose. In those times, human beings were both
manufacturers and users. This situation changed with time.
As civilisations arose, jobs were divided, and different people got specialised
in the production of different goods and services. For example, farmers
cultivated crops, potters made pots; weavers wove the cloth required for
everyone. This specialisation created a separation between the producer and
the user. Thus, the producer/manufacturer and the user were no longer the
same individuals. However, the goods were not so complicated and the user
had long familiarity with the products through prior use. Therefore, the user
could assure that the products were fit for use. The situation changed
drastically with the advent of industrialisation and the notion of quality control
underwent a sea change.
For the sake of interest, we describe below the important stages in the
historical evolution of quality control:
1. Operator or craftsman quality control
From early civilisation until the Industrial Revolution, quality control was
ensured/observed either by a single worker (operator or craftsman) or a
very small number of workers who totally controlled the quality of their
work. This situation continued until the nineteenth century.
2. Foreman or supervisor control
Large factories were established after the Industrial Rrevolution to meet
the increasing demand of consumers. Therefore, the number of workers
increased in the factories. This gave rise to the need for supervisors who
guided, controlled and administered to other workers. Generally, in those
times, the supervisor was picked up from among the workers of the
factory who had a thorough knowledge of the work. Checking the quality
of a job done by workers was also the responsibility of the supervisor
functions. This situation continued until the First World War.

3. Inspector of quality control

The manufacturing system became more complex due to

interchangeable components on a mass scale and it became essential to
critically examine each component with the help of measuring
instruments. Obviously, this job could not be performed by the
supervisor because he/she had expertise only in one component of the
product. Thus, a full-time inspector came on the scene to inspect the
finished goods and quality control was separated from production. 15
 Block 1 Process Control
4. Statistical quality control

Initially, quality control inspectors inspected all products. However, this

was costly and time-consuming. In 1924, Walter A. Shewhart, a
researcher at Bell Telephone Laboratories (a Research and
Development Unit of the American Telephone and Telegraph Company)
developed the technique of statistical control charts for the control of
product variability. This is usually considered the beginning of statistical
quality control (SQC).

In the same decade, Harold F. Dodge and Harry G. Roming, researchers

of Bell Telephone Laboratories, developed statistically based
Walter Andrew acceptance sampling plans as an alternative to 100% inspection.
Shewhart
(1891-1967) The techniques of control chart and acceptance sampling were used in
An American physicist, the bell system, but neither of these was widely adopted outside it.
engineer However, this changed during the Second World War. The ordinance
and statistician,
sometimes known as department of the U.S. Army was facing the problem of how to get large
the father of statistical quantities of arms and ammunition from multiple suppliers at acceptable
quality control.
levels of quality. In 1942, a quality control section was established in the
War Department and acceptance sampling came to be used widely.

After the Second World War, Japan was trying to revive itself from the
devastation of the world war. Japanese industries were almost
destroyed, and its leaders knew that rebuilding the industry was
essential for the survival of the nation. Japanese industrialists decided to
engage an American statistician as their consultant. The man they chose
was W. Edwards Deming, who had studied with Walter Shewhart.
Deming developed theories on how statistics could be used to improve
industrial quality. He gave a series of lectures in Japan on statistical
William Edwards Deming quality improvement methods. Japan adopted these methods and within
(1900-1993)
An American engineer, a few decades, became one of the most successful industrial nations on
statistician, professor, the earth.
author, lecturer and
management consultant.
So far, we have given you an overview of the concepts of quality and quality
control. We now discuss statistical quality control.

1.3 STATISTICAL QUALITY CONTROL

In the highly competitive market today, the main objective of manufacturers or
producers is to produce products of high quality and fulfil the specifications so
that they achieve quality assurance in manufacturing and service
organisations. To achieve this objective, different statistical tools have been
developed, which are useful for controlling the quality of products vis-a-vis the
specifications or standards. The technique of using statistical tools for
controlling product quality vis-a-vis specifications is known as Statistical
Quality Control (SQC).

Statistical quality control is defined as the technique of applying

statistical methods based on the theory of probability and sampling to
establish the quality standard and maintain it in the most economical
16 manner.
Unit 1 Overview of Statistical Quality Control
Let us now outline the elements that constitute SQC.

1.3.1 Elements of SQC

The following are the main elements of SQC:

(a) Sample Inspection

We know that 100% inspection needs huge expenditure of time, money,

labour and resources. Further, if the nature of the product is such that it
is completely destroyed during the process of inspection, e.g., a bulb,
candle, ammunition, food, etc., then 100% inspection is not practicable.
Therefore, SQC is based on sampling inspection. In the sampling
inspection method, some items or units (called sample) are randomly
selected from the process and then each and every unit of the sample is
inspected.

(b) Use of Statistical Methods

Some commonly used statistical tools such as random sampling, mean,

range, standard deviation, mean deviation, standard error and concepts
such as probability, binomial distribution, Poisson distribution, normal
distribution, etc., are used in SQC. Since the quality control method
involves extensive use of statistics, it is termed as Statistical Quality
Control.

(c) Fundamental Objective

The fundamental objective of SQC is to decide whether a unit produced

is according to its specifications or not. If the unit produced is not
according to its specifications and there is a variation in quality, it
becomes necessary to trace the causes of variation and eliminate them
if possible.

(d) Decision Making

With the help of statistical quality control, we decide whether the quality
of a product or a process of manufacturing/producing products/goods is
under control or not.

(e) Specifications, Production and Inspection

SQC method helps in deciding about the specifications, production and

inspection of a product.

We now describe the techniques of statistical quality control.

1.3.2 Techniques of SQC

The important techniques used for statistical quality control can be broadly
classified into two categories:

• Statistical Process Control or simply Process Control, and

• Product Control.

These techniques are further classified into different categories as shown in

Fig. 1.1. 17
Block 1 Process Control

Fig. 1.1: Classification of SQC techniques.

Let us discuss both categories of SQC techniques in some detail.

1.3.3 Statistical Process Control
Statistical Process Control or simply process control is the first part of SQC.
For understanding statistical process control, first of all, we should understand
the concept of process in quality control.
A process is a series of operations or actions that transforms input to
output. It is said to be stable or repeatable if the resulting output product is of
the given specifications or standard quality. But sometimes, due to certain
causes such as poor quality of raw material, changes in the machine settings,
use of the unskilled workforce, improper machines, etc., the stable process is
disturbed. In such situations, we require a tool or technique through which we
can control the process. This technique is known as statistical process
control (SPC).
Statistical process control is a technique used for understanding and
monitoring the process by collecting data on quality characteristics periodically
from the process, analysing them and taking suitable actions whenever there
is a difference between actual quality and the specifications or standard.
The statistical process control technique is widely used in almost all
manufacturing processes for achieving process stability and making
continuous improvements in product quality. Its major tools are:
1. Histogram
2. Check sheet
3. Pareto chart
4. Cause and effect diagram
5. Process flow diagram
6. Scatter diagram
7. Control chart
Of these seven tools, we shall describe only the technique of “control chart”
given by W.A. Shewhart in 1924 because it is the most preferred technique
18
today. In fact, it is probably an outstanding technique for controlling and
Unit 1 Overview of Statistical Quality Control
improving quality. We introduce the basics of control charts in Sec. 1.5. In
Units 2, 3 and 4 of this course, we discuss various types of control charts.
1.3.4 Product Control
In many situations, the product is so complex that a manufacturer cannot
make all components or parts of the product. Therefore, one or more
component (s) of a product are purchased from outside agents or suppliers
and the manufacturer does not have direct control over the quality of such
component (s). In such situations, the manufacturer faces the problem of
controlling the quality of component (s) sourced from elsewhere. In addition,
the manufacturer needs to control the quality of the final product. So
he/she also faces the problem of ensuring that the final product meets its
specifications and that various lots of the product do not contain an
excessively large number of defective items. Such types of problems come
under the category of product control.
Product control means controlling the products in such a way that they
are free from defects and conform to their specifications.
Initially, product control was done by 100% inspection, which means that each
and every unit produced or received from outside suppliers was inspected.
This type of inspection has the advantage of giving complete assurance that
all defective units have been eliminated from the inspected lot. However, it is
time-consuming and costly. Also, if a product is destroyed under inspection,
e.g., light bulb, crackers, ammunition, picture tube of the TV, etc., 100%
inspection is not possible.
Sampling inspection or acceptance sampling was developed as an alternative
to 100% inspection.
Acceptance sampling is a technique in which a small part or fraction of
items/units is selected randomly from a lot and the selected items/units are
inspected to decide whether the lot should be accepted or rejected on the
basis of the information obtained from sample inspection. This is how product
control is achieved through acceptance sampling.
Acceptance sampling is also of different types. We shall discuss it in detail in
Unit 6.
You may like to pause now and check your understanding of the various
aspects of SQC discussed in this section. Answer the following Self
Assessment Question.

SAQ 2
Choose the correct option from the following:
(i) Statistical quality control is a technique of
(a) process control
(b) product control
(c) both (a) and (b)
(d) neither (a) and (b)
(ii) The statistical techniques used in statistical quality control are 19
Block 1 Process Control
(a) control charts
(b) acceptance sampling plans
(c) both (a) and (b)
(d) neither (a) nor (b)

You have learnt that SQC techniques are used for process control and product
control so that the products of the desired quality may be manufactured. We
now explain the causes of variation in the process.

1.4 CAUSES OF VARIATION

Variation in manufactured products is inevitable; it is a fact of nature and
industrial life. Even when a production process is well-designed or carefully
maintained, no two products are identical. The difference between any two
products could be very large, moderate, very small or even undetectable
depending on the sources of variation. For example, the weight of a particular
model of automobile varies from unit to unit, the weight of packets of milk may
differ very slightly from each other, the length of refills of ball pens, the
diameter of cricket balls may also be different and so on. The existence of
variation in products affects quality. So the aim of SQC is to trace the sources
of such variation and try to eliminate them as far as possible. The causes of
variation are broadly classified into two categories:
1. Chance causes, and
2. Assignable causes.
1.4.1 Chance Causes
Chance causes are also known as random or natural or common causes.
Even in a well-designed or carefully maintained production process, variability
exists in the product due to some natural/random causes. Even if the process
is operated under the same conditions, that is, the quality of raw materials
used is the same and there is no change in the machine settings, operators or
the environment, there is a specific pattern of variability in the product. For
example, the diameter of ball bearings varies slightly, there is a slight variation
in the weight of cricket balls, the fuel efficiency of a particular model of
automobile varies slightly and so on. Such variability is due to different
common or chance causes, which may affect the process output in minor
ways. Such causes are known as chance causes of variation. These may
arise due to, the inflexibility of aged machines, variability in purchased
material, poor lighting, the extent of worker training or other non-obvious
reasons. These may or may not be present at the same time, but when taken
together produce random results. If the quality of the output varies too much
due to chance causes, the process must be redesigned or modified to
eliminate one or more of these causes. Since process redesigning or
modification is the responsibility of the management, the elimination of
common/chance causes of variation is usually the responsibility of the
management and not that of workers. It may not be possible to eliminate all
chance causes in a process. However, even if variation due to chance
causes is present in the production process, it is still said to be under
20 statistical control.
Unit 1 Overview of Statistical Quality Control
1.4.2 Assignable Causes
Another kind of variability may be present occasionally in the output of a
process. The causes of such type of variability are not due to the process
design, but take place because of changes in raw material, machine, operator,
environment or any other component of the process. These causes are called
assignable causes and are also known as special or non-random or
unnatural causes. Accidental improper setting of the machine, a worker
falling ill and still continuing to work, change of operators or shift, breakages,
misreading of scales, the batch of defective raw material, etc. are examples of
assignable causes.
Since the effect of assignable causes is localised within a process, these may
be eliminated by workers or their immediate supervisor. The variability due to
assignable causes is generally larger than the variability due to chance causes
and it usually represents an unacceptable level of process performance. A
process that is operating in the presence of assignable causes is said to be an
out-of-control process.
You can now check your understanding of the causes of variation by
answering the following Self Assessment Question.

SAQ 3
A company manufactures cricket balls. The statistical quality controller of the
company finds that there is a variation in the weight of cricket balls. Answers
the following:
(i) The variation in the weight of cricket balls may be due to
(a) assignable causes
(b) chance causes
(c) both (a) and (b)
(d) neither (a) nor (b)
(ii) If the variation in the weight is due to chance causes, it is
(a) controllable
(b) not controllable
(c) both (a) and (b)
(d) neither (a) nor (b)

Having explained the causes of variations in the production process, we now

introduce the technique of control chart. It is the main technique used for
process control.

1.5 CONTROL CHARTS

You will agree that graphical representation is one of the most sensitive
statistical instruments. So we can represent the quality characteristic of the
output product such as weight, length, diameter, defects, etc. graphically to
understand, describe or monitor process variation. The idea of representing
quality characteristics graphically was first given by Walter A. Shewhart. He 21
Block 1 Process Control
invented control charts for industrial processes to distinguish acceptably
(chance) variation from assignable variation. He observed that with the help of
control charts, the occurrence of assignable causes of variation could be
detected quickly and corrective action could be taken to eliminate them.
A control chart is a two-dimensional graphical display of a quality
characteristic that has been measured or computed in terms of mean or other
statistics from samples and plotted against the sample number or time at
which the sample is taken from the process.
The concept of the control chart is based on the theory of sampling and
probability. In a control chart, a sample statistic of a quality characteristic such
as mean, range, proportion of defective units, etc. is taken along the Y-axis
and the sample number or time is taken along the X-axis. A control chart
consists of three horizontal lines, which are described below:
1. Centre Line (CL) – The centre line of a control chart represents the
value which can have three different interpretations depending on the
available data. First, it can be the average value of the quality
characteristic or the average of the plotted points. Second, it can be
a standard or reference value, based on representative prior data or an
aimed (targeted) value based on specifications. Third, it can be the
population parameter if that value is known. The centre line is usually
represented by a solid line.
2. Upper Control Line – The upper control line represents the upper
value of the variation in the quality characteristic. So this line is
called upper control limit (UCL). Usually, the UCL is shown by a
dotted line.
3. Lower Control Line – The lower control line represents the lower value
of the variation in the quality characteristic. So this line is called
lower control limit (LCL). Usually, the LCL is shown by a dotted line.
The UCL and LCL also have three interpretations depending on the available
data same as the centre line. These limits are obtained using the concept of
3σ (three-sigma) limits, which we shall describe in Sec. 1.6.
For the construction of control charts, we select samples of few units/items at
regular intervals from the process. Then we measure the quality characteristic,
e.g., length of foil, diameter of ball bearing, weight of cricket ball, etc. for each
unit of the selected samples or count the number of defects, number of
defective units, etc. for each selected sample. After that, we calculate the
statistic such as mean, standard deviation, range, proportion of defective units,
etc. for each sample. Then the values of the calculated statistic are plotted
against the sample number or time. The sample points on the graph may be
joined by a line. Joining the sample points by line segments is not compulsory
but by doing so, we can easily understand their sequence (pattern) over time.
If all sample points lie on or in between the upper and lower control limits, the
control chart indicates that the process is under statistical control. That is,
only chance causes are present in the process. No assignable cause is
present in the process. However, if one or more sample points lie outside the
22
control limits, the control chart alarms (indicates) that the process is not under
Unit 1 Overview of Statistical Quality Control
statistical control. Some assignable causes are present in the process. To
bring the process under statistical control, it is necessary to investigate the
assignable causes and take corrective action to eliminate them.
However, a control chart cannot tell us what is wrong with it. In fact, it is
the responsibility of the supervisor or quality control manager to find out what
has gone wrong.
Note: Sometimes the sample points may be inside the control limits but may
behave in a systematic or non-random manner. This is also an indication that
the process is out of control. For example, suppose 22 of the last 25 sample
points lie below the centre line but above the lower control line and only 3 of
these points lie above the centre line but below the upper control line. This
pattern indicates that the process is not under statistical control because it
is not random in appearance. We shall explore this aspect in more detail in
Sec 1.7.
You may like to try the following Self Assessment Question before studying
further.

SAQ 4
Choose the correct option from the following:
(i) Control chart is a
(a) one-dimensional chart.
(b) two-dimensional chart.
(c) three-dimensional chart.
(d) none of the above.
(ii) Control chart consists of
(a) one control line.
(b) two control lines.
(c) three control lines.
(d) four control lines.

In Sec. 1.5, you have learnt about the control chart, which contains a centre
line (CL), lower control line (LCL) and upper control line (UCL). We now
discuss how to obtain the centre line and control limits for a control chart. The
UCL and LCL are also called 3σ limits.

1.6 3σ LIMITS
The quality characteristic can be described by a probability distribution or a
frequency distribution. In most situations, a quality characteristic follows a
normal distribution or can be approximated by a normal distribution. You
have studied the normal distribution in Unit 11 of MST-012 entitled Probability
and Probability Distributions Theory and you know the probability that a
normally distributed random variable (X) lies between µ − 3σ and µ + 3σ is
0.9973 where µ and σ are the mean and the standard deviation of the random
variable (X). Thus, 23
Block 1 Process Control
P [μ − 3σ ≤ X ≤ μ + 3σ ] =0.9973

So the probability that the random variable X lies outside the limits μ ± 3σ is
1– 0.9973 = 0.0027, which is very small. It means that if we consider 100
samples, most probably 0.27 of these may fall outside the μ ± 3σ limits. So if
an observation falls outside the 3σ limits of 100 observations, it is logical to
suspect that something might have gone wrong. Therefore, the control limits
on a control chart are set up by using 3σ limits. The UCL and LCL of a
control chart are called 3σ limits of the chart. The question is: How do we
calculate 3σ limits?
Suppose M is a sample statistic (e.g., mean, range, proportion of defectives,
etc.) that measures some quality characteristic of interest. Further suppose
that μM and σM are the mean and standard error (standard deviation) of the
sample statistic M, respectively. Then the centre line and control limits for
controlling the quality characteristic are given by:
Centre line (CL) = μM ... (1)

Upper control limit (UCL)

= μM + 3σM ... (2)

Lower control limit (LCL)= μM − 3σM ... (3)

Fig. 1.2: 3σ limits.

The chart in Fig. 1.2 shows the centre line (CL), lower control limit (LCL) and
upper control limit (UCL). The UCL and LCL are set at the distance ± 3σM
from the centre line (μM). Note from Fig. 1.2 that the area covered between the
UCL ( =μM + 3σM ) and LCL ( = μM + 3σM ) is 0.9973 (99.73%). So the probability
that an observation falls outside these limits is 0.0027.

If the sample points fall between the control lines, the process is said to be
under statistical control. But, if one or more points lie outside the control
limits, the control chart alarms (indicates) that the process is not under
statistical control. Some assignable causes are present in the process. To
bring the process under statistical control, it is necessary to investigate the
assignable causes and take corrective action to eliminate them and then
24 continue the production process.
Unit 1 Overview of Statistical Quality Control
You may now like to check your understanding of 3σ limits by answering the
following Self Assessment Question.

SAQ 5
If μ and σ represent the mean and standard deviation of the process, the lower
and upper three-sigma control limits for a control chart are given by:
(a) μ − 3σ 2 and μ + 3σ 2
(b) μ − 3σ and μ + 3σ

(c) μ2 − 3σ 2 and μ2 + 3σ 2

(d) μ2 − 3σ and μ2 + 3σ

In Sec. 1.6., you have learnt that the points on the control chart describe a
pattern. If one or more sample points fall outside the control limits, the process
is said to be out-of-control and if all sample points are inside or on the control
limits, the process is said to be under statistical control. But in many cases, the
sample points may lie within the control limits and yet show an unnatural or
specific pattern. This is also an indication of assignable causes. So to decide
whether the process is under statistical control or not, it is also important to
analyse the pattern of the sample points. This is what we discuss in
Sec. 1.7.

1.7 CONTROL CHART PATTERNS

The patterns of the control chart are broadly classified into two categories:
1. Natural patterns of variation, and
2. Unnatural patterns of variation.
We first discuss the natural patterns of variation.
1.7.1 Natural Patterns of Variation
You know that the central tendency is the characteristic of a distribution. Most
observations tend to concentrate near the centre (mean) of the distribution and
very few points lie near the tails. Since the normal distribution is symmetrical
about its mean (μ), i.e., the centre line is at μ (CL = µ ), we expect that half the
points will lie above the centre line and half below it. We also know that for the
normal distribution:
P [μ − 3σ ≤ X ≤ μ + 3σ ] =0.9973

This means that 99.73% of observations lie between the 3σ limits. So of a

total of 100 observations, 99.73 observations will lie inside the 3σ limits and
only 0.27 observations may lie outside the 3σ limits.
We may conclude that a control chart having a natural pattern of variation has
the following three characteristics:

(i) Most points lie near the centre line of the chart.

(ii) Very few points lie near the control limits.

(iii) None of the points falls outside the control limits. 25

Block 1 Process Control
A typical control chart with 3σ limits is shown in Fig. 1.3.

UCL

CL

LCL

Sample Number

Fig. 1.3: Typical control chart.

1.7.2 Unnatural Patterns of Variation

The Western Electric Company engineers have identified 15 unnatural typical
patterns in control charts. In this section, we discuss the most important types
of unnatural patterns.
1. Extreme Variation
If one or more points is/are significantly different from the other points
and lie outside the control limits of the control chart as shown in Fig.1.4,
we say that there is an extreme variation in the chart. Some assignable
causes are present in the process and corrective action is necessary to
bring the process under control.

UCL

CL

LCL

Sample Number

Fig. 1.4: Extreme variation.

Causes of extreme variation are:

(i) Error in measurement, recording and calculations,

(ii) Wrong setting and defective machine tools or erroneous use,

(iii) Power failures for short time,

26 (iv) Use of a new tool, failure of the component at the time of test, etc.
Unit 1 Overview of Statistical Quality Control
A sample point falling outside the control limits is a clear indication of the
presence of assignable causes. There are other situations where the
pattern of sample points on the chart indicates the presence of
assignable causes, although all points may lie within the control limits.
Such situations are discussed below:
2. Trend
If consecutive points on the control chart tend to move upward or
downward as shown in Fig. 1.5, it can be assumed that the process
indicates a trend. If proper care or corrective action is not taken, the
process may go out of control.

UCL

CL

LCL

Sample Number

Fig. 1.5: Trend pattern.

The causes of the trend pattern are:

(i) Tool or die wear,
(ii) The gradual change in temperature or humidity,
(iii) The gradual wearing of operating machine parts,
(iv) Gradual deterioration of equipment, etc.
3. Cycles
When consecutive points exhibit a cyclic pattern (as shown in Fig. 1.6), it
is also an indication that assignable causes are present in the process
which affect it periodically.

UCL

CL

LCL

Sample Number

Fig. 1.6: Cyclic pattern. 27

Block 1 Process Control
The causes of the cyclic pattern are:
(i) Rotation of operators,
(ii) Periodic changes in temperature and humidity,
(iii) Periodicity in the mechanical or chemical properties of the material,
(iv) Seasonal variation of incoming components, etc.
4. Shifts
When a series of consecutive points falls above or below the centre line
of the chart, it can be assumed that a shift in the process has taken
place. This indicates the presence of some assignable causes. Such a
pattern is shown in Fig. 1.7.

UCL

CL

LCL

Sample Number

Fig. 1.7: Shift pattern.

Generally, it is assumed that when 7 consecutive points lie above or

below the centre line, a shift has occurred.
The causes of the shift pattern are:

(i) Change in material,

(ii) Change in machine setting,

(iii) Change in operator, inspector, inspection equipment, etc.

5. Erratic Fluctuations

When the sample points of the control chart tend to fall near or slightly
outside the control limits with relatively few points near the centre line as
shown in Fig. 1.8, it can be assumed that erratic fluctuation has taken
place. This indicates the presence of some assignable causes. The
causes for erratic fluctuations are slightly difficult to identify. These may
be due to different causes acting at different times in the process.

The causes of erratic fluctuations are:

(i) Different types of materials being processed,

(ii) Change in operator, machine, inspection, equipment,

28 (iii) Frequent adjustment of the machine, etc.

Unit 1 Overview of Statistical Quality Control

UCL

CL

LCL

Sample Number

Fig. 1.8: Erratic fluctuations.

You may now like to review what you have studied about control chart
patterns.

SAQ 6
Choose the correct option from the following:
(i) If the points on the chart have continuous movements upward/downward,
the pattern is called
(a) freak pattern
(b) shift pattern
(c) trend pattern
(d) cyclic pattern
(ii) If all points on the control chart are within the control limits and the pattern
of the points shows trend, the process is said to be
(a) under control
(b) out-of-control
(c) both (a) and (b)
(d) neither (a) nor (b)

1.8 ADVANTAGES AND LIMITATIONS OF SQC

When a lot of items/units is manufactured, the manufacturer has two methods
to check the quality of the lot: firstly, he/she could check each and every item A lot is the collection
of units or items which
and decide about the quality of the product, i.e., 100% inspection. Secondly, are produced by the
he/she could use statistical quality control methods, i.e., inspect a small same machine, same
number of items and decide about the quality of an entire lot of the produced operators, using the
product. SQC has many advantages over 100% inspections, which are listed same raw materials
and approximately
below: during the same time
1. Ease of Application period.

An excellent feature of statistical quality control is that it is easy to apply.

While developing the statistical methods for quality control, skilled and 29
Block 1 Process Control
intelligent persons are required. However, even those persons who
have not had extensive specialised training can apply statistical methods
easily.
2. Reduction in Costs
The cost of the inspection is reduced. In SQC as only a part or fraction of
a lot is taken and inspected.
3. Greater Efficiency
Inspection of every item is bound to reduce the efficiency of quality
control inspectors because of dullness. Inspectors are more alert while
using SQC as only a part is inspected.
4. Early Detection of Faulty Units
SQC consists of continuous checking of the quality of the product. When
a sample point falls outside the control limits it gives the signal that the
process is not under statistical control. If some assignable causes are
present in the process, necessary corrective action can be taken.
Therefore, SQC ensures early detection of faults and results in minimum
wastage of items.
5. Helpful in Specification
Using SQC, we can find out whether the produced item is under control
or not, that is, whether the item meets the specifications within the
tolerance limits or not. If the variation is beyond the tolerance limits, SQC
gives a danger signal and necessary corrective action can be taken. So,
as long as statistical control continues, specifications can be accurately
predicted for the future, which cannot be guaranteed by 100%
inspection.
6. Ensures Overall Coordination
SQC methods ensure coordination between managers managing
specifications, production and inspection. It provides a basis to resolve
the differences arising among the various interests in an organisation.
7. Determination of the Effect of Change on the Process
With the help of control charts, we can easily detect whether or not a
change in the production process results in a significant change in
quality.
8. Equilibrium in Consumer’s and Producer’s Risk
Methods such as quality control and acceptance sampling help in
maintaining equilibrium between the consumer’s risk and the producer’s
risk.
9. Wider Applications
It is not only useful in the examination of items produced in small
numbers but also when bulk production has to be done.
10. Unique Method
Statistical quality control is helpful for those items that get destroyed on
being examined for a given quality characteristic, for example, the
30
intensity of match sticks, the average life of compact fluorescent lamp
Unit 1 Overview of Statistical Quality Control
(CFL), strength of glass, etc., In such cases, 100% inspection will spoil
the entire lot and create a huge loss.

However, SQC has some limitations, which are described below:

1. When a sample of the items, drawn from a lot, is not a true representative
of the entire lot (does not have the same characteristics as the lot from
which it is drawn), then a good lot may be rejected and a bad one may be
accepted. This is the main limitation of SQC.

2. SQC cannot be used mechanically for any production process without

studying the process and without adequate knowledge about it.

3. SQC applied to a production process provides only the information that

the process is under control or out of control. It cannot take any action for
improvement.

We end this unit by giving a summary of its contents.

1.9 SUMMARY
• Quality means:
 conforming to specifications,
 fitness for use,
 customer satisfaction,
 delighting the customer, and
 enchanting the customers.
• Quality has eight dimensions:
(i) Performance (ii) Features (iii) Reliability
(iv) Conformance (v) Durability (vi) Serviceability
(vii) Aesthetics (viii) Reputation
• Quality control is defined as the process by which we measure the
quality characteristics of the product, compare them with the
specifications or standard and take suitable actions whenever there is a
difference between actual quality and the specifications or standard.
• The technique of controlling product quality vis-a-vis the specifications
using statistical tools is known as Statistical Quality Control (SQC).
• Statistical quality control can be broadly classified into two categories:
• Statistical process control or simply process control, and
• Product control or acceptance sampling.
• Statistical process control is a technique used for understanding and
monitoring the process by collecting the data on quality characteristics
periodically from the process, analysing them and taking suitable actions
whenever there is a difference between actual quality and the
specifications.
31
Block 1 Process Control
• Product control means controlling the products in such a way that they
are free from defects and conform to their specifications.
• The causes of variation are broadly classified into two categories:
 Chance or random or natural or common causes, and
 Assignable or non-random or unnatural or special causes.
• The control chart is a two-dimensional graphical display of a quality
characteristic that has been measured or computed in terms of means or
other statistics from a sample and plotted against the sample number or
time at which the sample is taken from the process.
• A control chart consists of three horizontal lines:
Centre Line (CL) – The centre line of a control chart represents the
value which can have three different interpretations depending on the
available data. First, it can be the average value of the quality
characteristic or the average of the plotted points. Second, it can be
a standard or reference value, based on representative prior data or an
aimed (targeted) value based on specifications. Third, it can be the
population parameter if that value is known. The centre line is usually
represented by a solid line.
Upper Control Line – The upper control line represents the upper
value of the variation in the quality characteristic. So this line is
called the upper control limit (UCL). Usually, the UCL is shown by a
dotted line.
Lower Control Line – The lower control line represents the lower value
of the variation in the quality characteristic. So this line is called the
lower control limit (LCL). Usually, the LCL is shown by a dotted line.
• If all sample points lie on or in between the upper and lower control
limits, the control chart indicates that the process is under statistical
control. However, if one or more points lie outside the control limits,
the control chart alarms (indicates) that the process is not under
statistical control. Some assignable causes are present in the process.
• There are two types of patterns of sample points on the control chart:
 natural patterns of variation

 unnatural patterns of variation

1.10 TERMINAL QUESTIONS

Choose the correct option from the following:

1. Durability refers to a measure of

(a) product life

(b) specifications

(c) past performance of the product

32 (d) the probability of a product’s failure within a specified time

Unit 1 Overview of Statistical Quality Control
2. Process control is achieved through
(a) control charts
(b) acceptance sampling plans
(c) both (a) and (b)
(d) Neither (a) nor (b)
3. Product control is achieved through
(a) control charts
(b) acceptance sampling plans
(c) both (a) and (b)
(d) neither (a) nor (b)
4. If the variation in the weight of cricket balls is due to assignable causes, it
is
(a) controllable
(b) not controllable
(c) both (a) and (b)
(d) neither (a) nor (b)
5. The variation due to chance causes
(a) is tolerable
(b) does not affect the quality of a product
(c) is uncontrollable
(d) all of the above
6. The variation due to assignable causes
(a) can be removed
(b) cannot be removed
(c) can be removed sometimes
(d) can be removed most of the time
7. If one or more sample points lie outside the control limits, the control
chart indicates that
(a) there is no assignable cause in the process and the process is
under statistical control.
(b) there is at least one assignable cause in the process and the
process is under statistical control.
(c) there is at least one assignable cause in the process and the
process is out of statistical control.
(d) there is no assignable cause in the process and the process is out
of statistical control.
8. Control charts in statistical quality control are used for
(a) describing the pattern of variation. 33
Block 1 Process Control
(b) checking whether the variability in the product is within the
tolerance limits or not.
(c) both (a) and (b)
(d) neither (a) nor (b)
9. To check whether the process is under control or out of control, we see
(a) the pattern of the sample points on the chart
(b) the position of the sample points on the control chart
(c) both (a) and (b)
(d) neither (a) nor (b)

1.11 SOLUTIONS/ANSWERS
Self Assessment Questions (SAQs)
1. (i) Option (d) is the correct option because we know that quality has
to incorporate the following:
• conforming to specifications,
• fitness for use,
• customer satisfaction,
• delighting the customer, and
• enchanting the customers.
(ii) Option (c) is the correct option because we know that
performance is the primary operating characteristic of a product.
Features are the additional characteristics available in the product
along with the primary operating characteristics. Reliability refers
to the probability of a product’s failure within a specified time
period. Aesthetics means how a product looks, sounds, feels, etc.
2. (i) Option (c) is the correct option because we know that statistical
quality control is a technique of both process control and product
control.

(ii) Option (c) is the correct option because we know that statistical
quality control is a technique of both process control and product
control. The control charts are used for process control and
acceptance sampling plans are used for product control.

3. (i) Option (c) is the correct option because we know that the variation
in quality characteristics may be due to assignable causes and
chance causes.

(ii) Option (b) is the correct option because we know that the variation
due to chance causes is not controllable whereas the variation due
to assignable causes is controllable.

4. (i) Option (b) is the correct option because we know that the control
chart is a two-dimensional graphical display of a quality
34 characteristic.
Unit 1 Overview of Statistical Quality Control
(ii) Option (c) is the correct option because we know that the control
chart consists of the centre line, upper control line, and lower
control line.
5. Option (b) is the correct option because we know that the lower and
upper three sigma control limits for a control chart are μ ± 3σ where μ
and σ are the mean and the standard deviation of a process.
6. (i) Option (c) is the correct option because we know that if some
consecutive points on the control chart tend to move upward or
downward, it is called a trend pattern. If one sample point is
significantly different from the other points and lies outside the
control limits of the control chart, we say that there is an extreme
variation (freak) in the chart. If a series of consecutive points fall
above or below the centre line of the chart, it can be assumed that
a shift in the process has taken place. If points on the chart have
peaks and troughs which repeat themselves, we say that there is a
cyclic pattern. These patterns are also an indication of the
assignable causes.
(ii) Option (b) is the correct option because we know that if the sample
points show a trend, it is also an indication of assignable causes
and the process is out-of-control.

Terminal Questions (TQs)

1. Option (a) is the correct option because we know that durability means
a measure of product life. Conformance means meeting the
specifications or standards. Reputation is related to the past
performance of the company. Reliability refers to the probability of a
product’s failure within a specified time period.
2. Option (a) is the correct option because we know that process control is
achieved through control charts, whereas product control is achieved
through acceptance sampling plans.
3. Option (b) is the correct option because we know that product control is
achieved through acceptance sampling plans, whereas process control
is achieved through control charts.
4. Option (a) is the correct option because we know that the variation due
to assignable causes is controllable whereas the variation due to chance
causes is not controllable.
5. Option (d) is the correct option because we know that chance causes
affect the process output in minor ways.

6. Option (a) is the correct option because we know that the variation due
to assignable causes can be removed always.

7. Option (c) is the correct option because we know that if one or more
points lie outside the control limits, the process is not under statistical
control. Some assignable causes are present in the process.

8. Option (c) is the correct option because we know that control charts in
statistical quality control are used for describing the pattern of variation 35
Block 1 Process Control
and checking whether the variability in the product is within the tolerance
limits or not.

9. Option (c) is the correct option because for checking whether the
process is under control or out of control, we see the pattern of the
sample points on the chart as well as the position of the sample points
on the control chart.

36