1

Statistical Quality Control

Introduction
Statistical Quality Control (SQC) is a technique, which is used for applying statistical
theory to industrial mass production for ensuring uniformity in the items as also for facilitating in
meeting the specification requirements. In any manufacturing process, some variation in quality,
as measured by a parameter (s), of items is expected as a result of chance.

Statistical quality control has its origin nearly 1920. Mr. Dodge H.E. of the Bell telephone
laboratories in USA, first used normal distributions in statistical methods, to develop control
charts for checking the quality of the products by tests on random samples of products
manufactured by a factory.

The main concept of any production process is to maintain good quality of the product
manufacturing or produced. Now a day the manufacturing process is interested to see that the
product confirms to contain specifications or standards which are not rejected by the customers.

It is observed that in all production processes, no two items produced even by a machine
are exactly alike. When an accurate measurement is taken, variations in quality of the items
(threading of screws, thickness of wires, etc.) are inevitable. Specifications recognize this fact
by giving tolerance range with in which measurement must fall for the items to be acceptable. A
direct method to ensure that all of the items or parts fall within the specification limits is to
inspect each and every item is costly and the time consuming and the defective items are traced
only after production.

Also 100% inspection cannot ensure that only satisfactory products will come out of the
factory. However, the use of statistical methods like control charts is much more effective in
controlling the manufacturing process ensuring more or less information quality of product at
lower cost of production.

The main objective of SQC is to find out i) the extent to which the products satisfies the
specifications ii) the limits of variations and iii) the causes of variations.

Def: Statistical quality control (SQC) is defined as the statistical methods applied for the
maintenances of uniform quality in a continuous flow of products manufactured by an industry.

This can also be defined as “an economic and effective system of maintaining and
improving the quality of output throughout the whole operating process of specification”,
predictors and inspection based on continuous testing of random samples.

Advantages of SQC
Statistical Quality Control has several advantages over 100% inspection of a system.
1. Only sample of output inspected, inspection costs are tremendously reduced.
2. It is easy to apply.
3. Inspection Works is considerably reduced, efficiency is increased.
 2

4. SQC ensures an early detection of faults

5. It ensures that a minimum waste of reject production.
6. This enables a process to be brought into and held in a state of statistical control.
7. It is preferred as the only course where 100% inspection cannot be carried out without
destroying all the products inspected.
8. It is used to detect whether or not a change in the production process results in a
significant change in quality.

Types of Quality Control

[

The control of quality of a product manufactured by a factory can be classified under two
heads. Viz., Process Control and Product Control
1. Process Control
2. Product Control (or Lot control)
Control Charts
Control chart is a most important tool of statistical Quality control. It is a simple and
most powerful graphical device developed by Dr. W. A. Shewhart in 1924.

The main feature of the Control Chart method is the drawing of conclusions about the production
process on the basis of samples drawn from the production process at regular intervals.

This graph consists of a central horizontal line corresponding to the average value of the
characteristic, together with upper and lower control limits between which the values of the
sample statistic should fall.

These charts are used to detect lack of control in any manufacturing process and to
indicate any future trouble that any creep in the process, when the control charts indicate that any
assignable cause is present, steps should be taken to locate the trouble and correct it before more
defective products are produced. If these charts indicate that only random variations are present
in the products, then no further inspection will be necessary.

We know that the Process Control is achieved through Control Charts where as Product
Control is accomplished through Sampling Inspection Plans. There are two principal types of
control charts namely, Control Charts for Variables, and Control Charts for Attributes. Variables
are quality characteristics that can be measured and expressed in numbers, namely length of a
screw, diameter of a wire, etc.

Attributes are usually refer to the classification of a quality characteristic into any one of
two classes, either conforming or not conforming to specifications, as in acceptance or rejection
by sampling inspection plans. Attributes may be judged either by the proportion of defectives or
by the number of defects per unit.

Causes of Variations in Quality Characteristics

The causes of variations in quality characteristics can be classified into two ways.
1. Assignable causes of variation (Assignable variation)
2. Chance causes of variation (chances variation)
 3

Control Charts for Variables

There are three commonly used control charts for variables:
i) x -Chart or Mean Chart
ii) R- Chart or Range Chart
iii)  - Chart or Standard Deviation Chart

General procedure to draw Control Chart for variables

The construction of control chart is based on a sequence of suitable sample statistics
(namely, x , R,  ) calculated from samples drawn at random from the products of the
manufacturing process.

3-  Control Limits
Let  be the quality-characteristic of an item manufactured.
Let x1, x 2, x3, . . . x n be a random sample of size ‘n’ drawn from the given process and let ‘t’ be a
statistic computed from the sample such that
E (t )   i and V (t )   i2
If the process is under control,  will remain more or less the same from sample to
sample and hence the variations in the values of t from sample to sample is due to chance causes.
If the statistic t is normally distributed with mean  i and variance  i2 , then by the area
distribution of normal curve we have
 t  i 
P  3   3 =0.9974
 i 
P i  3 i  t   i  3 i  =0.9974

P t   i  3 i  = 0.9974

P t  i  3 i  = 0.0026
This shows that if the process is under control, then the observed values of the statistic t
lie between the limits i  3 i and  i  3 i in 99.74% of the cases and they lie outside these
limits in 0.26% cases that is very small.
In this case,  i  3 i is called the Upper Control Limits (UCL) and i  3 i , the Lower
Control Limit (LCL).

Similarly, we can find 2-  control limits when observed values of ‘t’ lie between  i  2 i and
 i  2 i in 95.44% of the cases.

Thus if for any random sample, the computed value of ‘t’ lies within the 3-  limits. (i.e.
within  i  3 i ), then we say that the process is in state of statistical control and in this case the
process is operating under the influence of chance causes of variation only. But if the computed
value of ‘t’ lies outside the 3-  limits,  i  3 i , then it is an indication the lack of statistical
 4

control. In this case, some assignable causes are present in the process and they should be
traced, identified and eliminated.

A control chart consists of one vertical line and three horizontal lines.

Quality scale is marked according to the quality characteristic of each sample along the
vertical line. The central line representing the average quality of the samples is plotted as a bold
horizontal line CC  . Upper Control Limit (UCL) and Lower Control Limit (LCL) are plotted as
dotted horizontal lines AA and BB  respectively, above and below the central line CC  , each of
them being at a distance of 3 i from CC 

Sample Numbers are shown at equal distances on a horizontal line at the bottom of
the control chart.
In control chart, if all the plotted points corresponding to the given samples lie within the
control limits i.e. in the region between the two dotted lines AA and BB  , then the process is said
to be in a state of control. Otherwise the process is not in a state of control. In such a case,
production process should be stopped assignable causes present in the process must be traced,
identified and eliminated before the production starts again.
Before drawing a control chart, it is important to group the manufactured items into
suitable samples in such a manner that the variation in quality among the items within the same
sample is small, but between the items of any two samples is as large as possible. Such a sample
is called a rational sub-group. The variation in quality between any two samples is due to
chance causes, but appreciable variation within the same sample, if any, is due to assignable
causes which are to be identified and removed.

Control Chart for Mean or x -Chart

A control chart for Mean or x -Chart is the control chart on which the means of the
samples drawn from a production process at equal intervals of time are plotted as points. It
indicates the fluctuations of the means of the samples about the mean of the process. x -Chart is
used to determine whether these fluctuations are due to chance causes or assignable causes. The
construction of x -Chart involve the following steps.
i) Mean x i and Range Ri of samples drawn
 5

Suppose we draw k independent random samples each of size n (usually k lies between
20 and 30 and n = 4 or 5) from the lot of items to be examined. Let x1 , x 2 ,. . . , x k be the
sample means and R1 , R2 , . . . , Rk the sample ranges of the k samples. Then
sum of the observatio ns in the i th sample
xi 
n
And Ri  Li  S i = Largest observation – Smallest observation, in the ith sample (Since Range
Ri is the difference between the largest and smallest observations)
ii) Mean of the sample means x and mean of the sample ranges R :
The mean of all the k sample means is denoted by x and the mean of the all the k sample
ranges is denoted by R as follows:

x 
x1  x 2    x k
=
x i
and R =
R1  R2    Rk
=  i
R
k k k k
iii) Computation of Control Limits UCL x and LCL x
We assume that the quality characteristic X follows normal distribution with mean  and
standard deviation  . Here the process mean is  and the process standard deviation is 
The sample mean x approximately follows normal distribution with

E (x)   and the Standard Error of x , i.e., S .E ( x) 
n
The 3   control limits for x chart are
3 3
  3S .E ( x) =   =   A where A = , is a constant whose values for
n n
different values of n from 2 to 25 are given in table. The value of A are usually given in a
problem.
Case I: Standards are given: If  and  are given
i.e., If  =   and  =   are known then the central line is CL =  
And UCL x = Upper Control limit of x =   +A  
3
LCL x = Lower Control limit of x =   - A   where A =
n
Case II: Standards are not given: If  and  are not given
If  and  are not known then x is taken as the estimate of . We can estimate 
either with the help of the mean of the sample ranges R or on the basis of the means of the
sample standard deviations s and quality control factors as shown below.
 R
i) With the help of R ,  = the estimate of  = , where d 2 is a quality control factor
d2
depending on the size of the sample and it can be obtained from the Statistical Table.

In this case , the 3   control limits for x chart are

 6

3R 3
x = x  A2 R where A2 = is the constant ( A2 is quality control factor
d2 n d2 n
depending on the sample size n)

Now the central line is CL = x (If  and  are not given)

The Upper Control Limit UCL x = x  A2 R

The Lower Control Limit LCL x = x  A2 R

The values of A2 are n = 2 to n = 25 are given in table.
ii) With the help of mean of the sample standard deviations s :
  s
The estimate of  denoted by  and is taken as  = ,
c2
(where c2 is the quality control factor depending on sample size n.)
In this case , the 3   control limits for x chart are:
3s
x = x  A1 s
c2 n
3
where A1 = is the constant (it is also called quality control factor depending on the
c2 n
sample size n)
Now the Central Line is CL x = x
The Upper Control Limit UCL x = x  A1 s
The Lower Control Limit LCL x = x  A1 s
Construction of x chart:
To draw the x chart on a graph paper, we represent the sample numbers on a horizontal
scale at the bottom of the control chart and statistic x along the vertical scale. Sample means x 1 ,
x 2 ,. . . x k are then plotted as points against the k sample number 1, 2, . . . k. The central line
CL=   or
x , UCL x and LCL x are drawn as dark dotted lines at the computed values.
Comments on x chart:
If all the plotted points corresponding to the given samples fall within the control limits
i.e., in the region between the two dotted lines BB  and AA drawn at the UCL x and LCL x ,
then we say that the process is in a state of control.
But if one or more points fall outside the Upper and Lower lines i.e., above AA or BB 
below, then the process is not in a state of statistical control. In such a case the production
process is stopped, assignable causes present in the process is traced, identified and eliminated
before the production starts again

Control Chart for Range or R-Chart:

 7

R chart is used to control the variability in the quality of products manufactured by a

given process. The method of constructions of R-chart is similar to that of x - chart.
The 3-  control limits for R- chart are:
The Upper Control Limit UCLR = R +3  R
The Lower Control Limit LCLR = R - 3 R
Where  R is the standard deviation of all the k sample ranges R1, R2 .... .Rk, each of size ‘n’.

Remark: Control limits can also be determined for Range or R-Chart by using quality control
factors D1, D2, D3 and D4 whose values depends on the sample size ‘n’ as

i) If the process standard deviation  is known then the control limits are:
The Upper Control Limit UCLR = D2 
The Lower Control Limit LCLR = D1 
And the Central Line is CLR = d2 
ii) If the process standard deviation  is not known, then the control limits are:
The Upper Control Limit UCLR = D4 R
The Lower Control Limit LCLR =D3 R
And the Central Line is CLR= R

The sample numbers 1, 2, . . . k are represent along a horizontal line and the range statistic R
represent along the vertical line on a graph paper. Sample range R1, R2, . . . . Rk are then plotted
as points against the respective sample numbers 1, 2,…, k.
The Central Line CL at R , UCLR and LCLR at the computed values are drawn as dark dotted lines
CC  , AA & BB respectively as

Note: If all the points fall between the two control lines BB and AA then the variation in
quality of the manufactured products is in a state of control. Otherwise the process is not in a
state of statistical control.
Observation (Choice of x and R Chart) :
In order to examine whether a production process is in s state of control or not, it is
necessary to study both x and R charts together. x chart indicate the variations between the
 8

samples, where as R chart indicates variations within the samples. So a production process is
said to be under control if both the charts indicates that the process is in a state of control.
It is better to construct R chart first. If R chart shows that dispersion (or variation) of the
quality in the process is not in a state of control, it is better to construct x chart until the quality
dispersion is brought under control. If a point or a set of points fall outside the control limits in
x -chart, but R chart indicates a state of control, then we say that the process average has shifted
considerably.

Note:
In any control chart, the calculated Value of Lower Control Limit (LCL) becomes
negative, then we take it as zero, because neither x nor R can be negative.

Applications
Example 1: Construct X chart for the following data.
Sample Observations
Numbers I II III
1 32 36 42
2 28 32 40
3 39 52 28
4 50 42 31
5 42 45 34
6 50 29 21
7 44 52 35
8 22 35 44
Also determine whether the process is under control or not and prepare a control chart

Solution:-
From the given data, n=3 and k=8

Observations Sample
Sample Total Sample
mean
Numbers I II III x x
Range-R

1 32 37 42 111 37.00 42-32 = 10

2 28 32 40 100 33.33 40-28 = 12
3 39 52 28 119 39.67 52-28 = 24
4 50 42 31 123 41.00 50-31 = 19
5 42 45 34 121 40.33 42-34 = 11
6 50 29 21 100 33.33 50-21 = 29
7 44 52 35 131 43.67 52-35 = 17
8 22 35 44 101 33.67 44-22 = 22
Total 906 302.00  R  144
From the above table  x  302.00  R  144
The mean of the means,
 9

1 302.00
x
k
 xi =
8
= 37.75
The mean of the ranges,
1 144
R   Ri = = 18
k 8
(From the table of control chart, for the sample size n = 3, we have A2  1.023 )
The Central Line CLx  x = 37.75
The control limits for 𝑥̅ Chart are x  A2 R
i.e., The Control Limits for mean chart are x  A2 R = 37.75  1.023 18
Therefore the Upper Control Limit UCLx = x  A2 R
= 37.75  1.023 18  56.164
The Lower Control Limit LCL x = x  A2 R
= 37.75  1.023 18  19.336
Now plot these points on the following graph with LCL x , CLx and UCLx

Sample Number 1 2 3 4 5 6 7 8
Sample Mean x 37.00 33.33 39.67 41.00 40.33 33.33 43.67 33.67

x  or Mean Chart
Since all the sample means lie between the control limits, the process is in under control
Example 2: A plastic manufacturer extrudes blanks for use in the manufactures of eye-glass
temples. Specifications requires that the thickness of these blanks have x =0.150 and  =0.002
inch. Calculate 2-signa and 3-sigma upper and lower control limits for means of samples 5 and
prepare a control chart

Solution:
From the given data,
n  5 , x = 0.150 and  = 0.002
For 2   Control Limits are:
 10

 0.002
x2  0.150  2   0.150  0.0017888
n 5

The Upper Control Limit, UCLx  x  2  0.150  0.0017888 = 0.1517 inch
n
Therefore the Upper Control Limit (UCL) =0.1517 inch

The Lower Control Limit, LCL x = x  2  0.150  0.0017888 = 0.1517 inch
n
Therefore the Lower Control Limit (LCL) = 0.1482 inch
The Central Line CLx  x = 0.150 inch

For 3   Control Limits are:

 0.002
x3  0.150  3   0.150  0.00268
n 5

The Upper Control Limit, UCLx  x  3  0.150  0.00268 =0.1526
n
Therefore the Upper Control Limit (UCL) =0.1526 inch

The Lower Control Limit, LCL x = x  3  0.150  0.00268 = 0.14732
n
Therefore the Lower Control Limit (LCL) = 0.14732 inch
The Central Line CLx  x = 0.150 inch
 11

Example 3: A machine is set to deliver packets of a given weight. Ten samples of size 5 each
were recorded. Below are given relevant data
Sample Number 1 2 3 4 5 6 7 8 9 10
Range 7 7 4 9 8 7 12 4 11 5
Draw the R-chart and comment on its state of control.
Solution:
From the given data, ten samples of size 5 each were recorded, and their range were
Sample Number 1 2 3 4 5 6 7 8 9 10 Total
Range 7 7 4 9 8 7 12 4 11 5 74

The Control Limits for the Range- chart,

Since R  74  7.4
10
Then the Central Line is CL= R = 7.4;
And the control limits are D4 R and D3 R when 𝜎 is unknown ( D 3  0; D4  2.115 )
 The Upper Control Limit is UCLR = D4 R  2.115  7.4 = 15.651
The Lower Control Limit is LCLR = D3 R = 0  7.4 = 0

Plot the following points on the graph along with LCLR , R , LCLR

Sample Number 1 2 3 4 5 6 7 8 9 10 Total

Range 7 7 4 9 8 7 12 4 11 5 74

From the control chart for Range, we see that the plotted points corresponding to all the sample
numbers fall within the control limits and, therefore, the process is in a state of control,
regarding the weights of the deliver packets.
 12

Control charts for Attributes

There are three types of control charts for Attributes:
(i) p- chart or control chart for fraction defective (or proportion of defectives);
(ii) np-chart or control chart for the number of defectives;
(iii) c- chart or control chart for the number of defects per unit.

P- Chart or control chart for fraction defective (or proportion of defectives)

The Control Chart p, i.e., p-chart is used to see the variation in sample units of
manufacturing process is there are not (variation in yes or no type), i.e., there are only two
possible outcomes: either the item is defective or it is not defective.
A product or service is defective if it fails to conform to specifications or a standard in some
respect. For example, consider the case of a customer calling the company to place an order. The
customer would probably not like to have the phone ring 10 to 15 times before it is answered.
Suppose you have determined that the operational definition for answering the phone in a timely
fashion is "to answer the phone on three or fewer rings." Using this definition, you could monitor
the fraction of phone calls answered or not answered in a timely fashion. If a phone call is
answered on or before the third ring, the item (answering the phone call) is not defective. If the
phone call is not answered on or before the third ring, the item is defective.
You use a p control chart when you have yes/no type data. Let P be the fraction defective in a
production process. Suppose a random sample of n units is drawn from the process and let d be
the number of defective units in the sample. Then the fraction defective p in the sample is given
by
d
p  or d  np
n
Then for the sampling distribution of the statistic p, we have
PQ P(1  P)
E(p) = P and S.E. of p =  Q  1 P
n n
P(1  P)
The 3-𝜎 control limits for p-chart are = p  3S .E.( p )  p  3
n
Case I :- Standards are given
If P be the given value of P, then the control limits for p-chart are
P(1  P) ,
P  3
n
Now Therefore the Upper Control Limit for p is
i.e  UCL p  P  3 P(1  P)
n
And the Lower Control Limit LCLp  P  3 P(1  P)
n
The Central Line is CL p = P 
Case II :- Standards are not given
p1  p2    pk
If P is not given, then p is an unbiased estimate of P and p  where pi is the
k
number of defectives in the ith sample.
 13

p( 1  p ) ,
The control limits for p-chart are p3
n

Now therefore the Upper Lower Control Limit for p are

i.e., UCLp  p  3 p( 1  p )  p  A p( 1  p )
n

And LCLp = p  3 p( 1  p )  p  A p( 1  p )
n

Where A  3/ n is called quality control factor which is constant.

The Central Line is CL p = p

Construction of p- chart is exactly similar to that of x  chart and R-chart. In this case we
represent the fraction defectives p1, p2, … .pk of the k samples along the vertical scale and they
are plotted as points against the corresponding sample numbers 1,2, … k

np- chart or control Chart for the number of defectives

If the sample size n is constant, i.e., the number of units in all the samples (or Sub–
groups) is the same, then d is the number of defective in a sample is equal to np, where p is the
faction defective in the sample.
For the sampling distribution of the statistic d,
E (d )  E (np)  nP and
S.E (d )  S.E (np)  nPQ
where P is the process (population) fraction defective and Q 1  P
The 3 control limits for np-chart are
np  3SE(np) =  np  3 nPQ where Q 1  P
Case 1:- Standards are given
If P  be the given value of P, then the control limits for np-chart are
nP  3 nPQ ,
Now the Upper Control Limit for np is
i.e UCLnp  nP  3 nPQ
And the Lower Control Limit LCLnp  nP  3 nPQ
The Central Line is CL p = nP 
Case-II Standards not given
If P is not given or not known, then p is an unbiased estimate of P and p  d = d
n k
th
where pi is the number of defectives in the i sample.
The 3 control limits for np-chart are n p  3 n p( 1  p ) ,
Now the Upper and Lower Control Limit for np is
 14

i.e., UCLnp = n p  3 n p( 1  p )

And LCLnp = n p  3 n p( 1  p )
The Central Line is CL p = n p
Note: If the sample size n is constant for all the samples i.e. n1  n2    nk  n (say ), then
np-chart (or d-chart) should be used. But if the sample size varies from sample to sample, the p-
chart is more suitable to examine the state of control.
C-chart or control chart for the number of defects per unit
It is constructed to control the number of defects per unit of the products. Every defective
unit may contain one or more defects. If a variable c represents the number of defects observed
per unit (i.e per item), then c follows Poisson distribution with parameter, say m. Thus the 3-𝜎
control limits for c -chart are based on the Poisson distribution whose standard deviation is m
As the statistic c is distributed as a Poisson variable with parameter m , we have E (c)  m and
standard error of c is standard deviation of c is equal to m

Case I: Standards are given

If m  be the given value of m, then the 3 control limits for c-chart are
m  3 m ,
Now the Upper Lower Control Limits for c are
i.e UCL c  m  3 m And LCL c  m  3 m
The Central Line is CL p = m 
Case –II:- Standard are not given
If m is not given, then c is an unbiased estimate of m and

c1  c2    ck
mc
k
where c i is the number of defects observed in the ith item of the product.

The 3 control limits for c-chart are c  3 c ,

Now the Upper and Lower Control Limit for c are
i.e., UCL c  c  3 c And LCL c = c  3 c
The Central Line is CL p = c
Note: If in any case lower control limit comes out to be negative, we take it as zero, since LCL c
is always greater than or equal to zero.
 15

Applications
Example 1: 15 samples of 200 items each wave drawn from the output of a process. The
number of defective items in the samples is given below. Prepare a control charts for the fraction
defective and comment are the state of control.
Sample
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15
Number of
12 15 10 8 19 15 17 11 13 20 10 8 9 5 8
Defective (np)

Sol:
From the given data, n = 200 Since each sample contains 200 items; and k = 15
Sample
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 Total
Number of
12 15 10 8 19 15 17 11 13 20 10 8 9 5 8 180
Defective (np)

and from the above table  np  180

n p 
 np  180  12
k 15
1 1
 p  12   12  0.06
n 200
For the P-chart
The Central Line CLP = p =0.06

The control limits are P  3 P(1  P )

0.06(1  0.06)
 0.06  3
200
0.06  094
 0.06  3
200
0.0564
 0.06  3  0.06  3  0.01679
200
 0.06  0.05038
The upper control limits for p-chart is
P(1  P)
UCLP = P  3
n
= 0.06 + 0.05.38 = 0.11038
The lower control limits for p-chart is
 16

P(1  P)
LCLP
 P 3
n
= 0.06 - 0.05038 = 0.0096
The fraction defectives (values of p) for the given samples are : 0.06, 0.07, 0.05, 0.04, 0.095,
0.075, 0.085, 0.055, 0.065, 0.1, 0.05, 0.04, 0.045, 0.025, 0.04.
Since all the sample points lie between the LCL and UCL lines, the process is under control.
Remark: If the sub-group size n is not constant, then we find two sets of control limits for p-
chart, one set for the largest value of n and the other set for the smallest value of n. if all the
plotted points fall within the inner control limits corresponding to the smallest value of n, then
the process in statistically under control. But if one or more sample points fall outside the
control limits, then the process is out of control and action must be taken to locate the assignable
causes and to remove them. If one or more sample points fall between the two sets of limits,
then the actual control limits should be computed.
Example 2: In a factory producing spark plugs, the number of defectives found in the
inspection of 15 lots of 100 each is given below. Draw the control chart for the number of
defectives and comment on the state of control.
Sample:
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15
Number of
5 10 12 8 6 4 6 3 4 5 4 7 9 3 4
Defective (np)

Sol:
From the given data, n = 100; each sample contains 100 items, and k = 15,
Sample:
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 Total
Number of
5 10 12 8 6 4 6 3 4 5 4 7 9 3 4 90
Defective (np)

From the table,  np  90

1 1 1 1
Now n p =   np   90  6 and P = 6  6  0.06
k 15 n 100
for the np. Chart:
Central line CLnp = n P  6

The control limits for np chart are: n P  3 n P(1  P )

= 6  3 6  (1  0.06 ) = 6  3 6  0.94  6  3 5.64
= 6  3 2.3749 = 6  7.1247
The upper control limit for np chart is UCLnp  n P  3 n P(1  P)
= 6  7.1247 = 13.1247.
 17

The lower control limit for np chart in UCLnp  n P  3 n P (1  P)

= 6 - 7.1247 = -1.1247
Since LCLnp cannot be negative; UCLnp  0
Since al the sample points lie between the upper and lower control lines, the process in under
control.
Example 3: Ten pieces of cloth out of different rolls of equal length contained the following
number of defects: 3, 0, 2, 8, 4, 2, 1, 3, 7, 1. Prepare a c-chart and state whether the production
process is in a state of statistical control.
Sol.
If c be the number of defects in an item, then
c = 3, 0, 2, 8, 4, 2, 1, 3, 7, 1.

c 
c  3  0  2  8  4  2 1 3  7 1 
31
 3.1
k 10 10
The control limits for the c-chart are
UCLc  c  3 c  3.1  3 3.1  3.1  5.282  8.382
And LCL c = c  3 c  3.1  5.282  2.182
Here we take LCL c as zero
The central line is CLc = c  3.1

From the c-chart shown in Fig. we see that all the plotted-point fall between the upper and
lower control limits.
Hence the production process is in a state of statistical control.