3/28/2021

IPE 4205
Introduction to Acceptance Sampling
Kazi Wahadul Hasan
Ahsanullah University of Science & Tech. Slide 1 of 29 Quality Management

Introduction to Acceptance Sampling

Chapter Outline
 Introduction
 Types of sampling plans
(i) attribute type
(ii) variable type
 Acceptance Sampling by Attributes
(i) Single Sampling Plan
(ii) Double Sampling Plan
(iii) Sequential Sampling Plan
 Military and ANSI Standards

Ahsanullah University of Science & Tech. Slide 2 of 29 Quality Management

1
 3/28/2021

Acceptance Sampling
 Popularized by Dodge and Romig, Acceptance
sampling is concerned with inspection and decision
making regarding products.
 Acceptance Sampling:
1. Is a middle of the road approach
2. Is the basis of “Statistical Inference”

Ahsanullah University of Science & Tech. Slide 3 of 29 Quality Management

Acceptance Sampling
Three approaches to lot sentencing:
1. Accept with no inspection
2. 100% inspection
3. Acceptance sampling

Acceptance sampling is the middle of road approach

between no inspection and 100% inspection.

Ahsanullah University of Science & Tech. Slide 4 of 29 Quality Management

2
 3/28/2021

Acceptance Sampling

Why Acceptance Sampling and Not 100%

Inspection?
• Testing is destructive
• Cost of 100% inspection is high
• 100% inspection is not feasible (require too
much time, technology or efforts)
• When the population or lot size is very large
• If vendor has excellent quality history

Ahsanullah University of Science & Tech. Slide 5 of 29 Quality Management

Acceptance Sampling
Advantages and Disadvantages of Acceptance Sampling
Advantages
• Economical
• Requires less time and effort
• Requires fewer personnel
Disadvantages
• Risk of accepting “bad” lots, rejecting “good”
lots. (consumer’s risk and producer’s risk)
• Less information generated
• Requires expertise in statistical aspects

Ahsanullah University of Science & Tech. Slide 6 of 29 Quality Management

3
 3/28/2021

Types of Sampling Plans

Two Major types of acceptance sampling plans.

• Acceptance sampling by Attributes
• Acceptance sampling by Variables
Acceptance sampling plans by Attributes
1. Single sampling plan
2. Double-sampling plan
3. Multiple-sampling plan
4. Sequential-sampling

Ahsanullah University of Science & Tech. Slide 7 of 29 Quality Management

Terminologies Related to
Sampling Plans
• Acceptable Quality Level (AQL)
The poorest quality level for the supplier’s process that a consumer
would consider to be acceptable as a process average. It is also
considered as a percent defective that is the baseline requirement
for the quality of a product. The producer would like to design a
sampling plan such that there is a high probability of accepting a lot
that has a defect level less or equal to AQL.
• Lot Tolerance Percent Defective(LTPD)
The poorest quality that a consumer is willing to tolerate in an
individual lot. The LTPD is a designated defect level for a lot
beyond which the lot is unacceptable to the customer. LTPD is a
level of lot quality specified by the consumer, not a characteristic of
the sampling plan

Ahsanullah University of Science & Tech. Slide 8 of 29 Quality Management

4
 3/28/2021

Terminologies Related to
Sampling Plans
• Type I Error (Producer’s Risk)
Conclude the shipment is bad when in fact it is good (false
alarm). In detail, this is the probability for a given sampling
plan of rejecting a lot that has a defect level less or equal to
AQL. The symbol a is commonly used for the Type I error and
the range of values lie within from 0.2 to 0.01.
• Type II Error (Consumer’s Risk)
Conclude the shipment is good when in fact it is bad
(overlooked problem). In detail, this is the probability for a
given sampling plan of accepting a lot with a defect level equal
or more than the LTPD. The symbol b is commonly used for
the Type II error and the range of values lie within from 0.2 to
0.01.

Ahsanullah University of Science & Tech. Slide 9 of 29 Quality Management

Terminologies Related to
Sampling Plans
• Operating Characteristic (OC) Curve
 The operating-characteristic (OC) curve measures the performance
of an acceptance-sampling plan.
 The OC curve plots the probability of accepting the lot versus the lot
fraction nonconforming.
 The OC curve shows the probability that a lot submitted with a
certain fraction defective will be either accepted or rejected.

Ahsanullah University of Science & Tech. Slide 10 of 29 Quality Management

5
 3/28/2021

Terminologies Related to
Sampling Plans
• Average Outgoing Quality (AOQ)
In a non-destructive testing, rejected lots are inspected 100% and
the identified defective units are replaced with good units. In such
case, all rejected lots are made perfect and the only defects are
those in lots that were accepted. AOQs refer to the long term defect
level for this combined LASP and 100% inspection of rejected lots
process. Therefore, if all lots of size N come with a defect level of P,
( )
AOQ =

• Average Sample Number (ASN)

The average sample number (ASN) is the expected number of
observations required to make a lot disposition decision. It is
mainly calculated in case of using double, multiple or sequential
sampling plan.
Ahsanullah University of Science & Tech. Slide 11 of 29 Quality Management

Parameters and Symbols

N = Total number of products/units in a lot or batch
n = Number of products/units to be inspected from a batch
D = Number of nonconforming units in a sample of size n
c = Acceptance number, or the maximum allowable number
of nonconforming units in a sample of size n.
p = Fraction nonconforming
Pa = Probability of acceptance
a = Producer’s Risk (Type-I error)
b = Consumer’s Risk (Type-II error)

Ahsanullah University of Science & Tech. Slide 12 of 29 Quality Management

6
 3/28/2021

Attribute Acceptance Sampling

Single Sampling Plans
In single sampling, a sample (of size n) is drawn from a lot (of size N). If
the number of nonconforming units (D) is less than a predetermined
number (known as acceptance number, c), then the lot or batch is
accepted, otherwise rejected.
In a single sampling plan,
 Accept/reject decision is made based on a single trial.
 Since the acceptance is “go/no go” type, therefore, single sampling
acceptance plan follows binomial distribution.
 The OC curve plays a crucial role in single sampling plan.
The probability of acceptance,

Ahsanullah University of Science & Tech. Slide 13 of 29 Quality Management

Example Problem
Suppose we have a sampling plan defined by the following
parameters:
n = 100, c = 2
What is the probability of accepting a lot with 0.5%
defectives? Plot the OC curve for this sampling plan

𝑛!
𝑃 =𝑃 𝑘≤𝑐 = 𝑝 1 − 𝑝1)
𝑘! 𝑛 − 𝑘 !

100!
= 0.0050) 1 − 0.0050)
𝑘! 100 − 𝑘 !

Ahsanullah University of Science & Tech. Slide 14 of 29 Quality Management

7
 3/28/2021

Example (Constructing p chart)

100!
= 0.0050) 1 − 0.0050)
𝑘! 100 − 𝑘 !

100!
= 0.0050) (1 − 0.0050) +
0! 100 − 0 !
100!
0.0050) (1 − 0.0050) +
1! 100 − 1 !
100!
0.0050) 1 − 0.0050)
2! 100 − 2 !
= 0.6058 + 0.3044 + 0.0757 = 0.9859

Similarly, calculate 𝑃 for other different p such as 0.01,

0.03, 0.05, 0.07, 0.09, 0.11, 0.15 etc. and plot the OC
curve from the output.

Ahsanullah University of Science & Tech. Slide 15 of 29 Quality Management

Double Sampling Plans

Ahsanullah University of Science & Tech. Slide 16 of 29 Quality Management

8
 3/28/2021

Double Sampling Plan

Ahsanullah University of Science & Tech. Slide 17 of 29 Quality Management

Example
A lot of 3500 books arrives at a bookstore. To check if the entire lot is
acceptable, a first random sample of n1 = 150 books is taken from the
lot and if c1 = 1 or less of the books are found to be defective, the entire
lot is accepted; if greater than c2 = 5 books are found, the lot is
rejected. If between 2 and 5 books are found to be defective on the
first sample, a second sample is taken of n2 = 100 books and if the
combined number of defects of the first and second samples is less or
equal to c2 = 5, the lot is accepted or rejected otherwise. Find the total
probability of acceptance at p = 0.03.

Ahsanullah University of Science & Tech. Slide 18 of 29 Quality Management

9
 3/28/2021

Solution
Calculating the probability of acceptance in the first trial:
𝑛 !
𝑃 = 𝑑 ≤𝑐 = 𝑝 (1 − 𝑝)
𝑑 ! 𝑛 −𝑑 !

150!
= 𝑝 (1 − 𝑝)
𝑑 ! (150 − 𝑑 )!

Calculating the probability of acceptance in the second trial:

𝑃 = 𝑑 +𝑑 ≤𝑐

𝑛 ! 𝑛 !
= 𝑝 (1 − 𝑝) ( 𝑝 (1 − 𝑝)
𝑑 ! 𝑛 −𝑑 ! 𝑑 ! 𝑛 −𝑑 !

150! 100!
= 𝑝 1 − 𝑝 1) 𝑝 1 − 𝑝1)
𝑑 ! 150 − 𝑑 ! 𝑑 ! 100 − 𝑑 !

Ahsanullah University of Science & Tech. Slide 19 of 29 Quality Management

Solution (Cont.)
d1 d2 ( 0 to c2-d1)
2 0,1,2,3
3 0,1,2
4 0,1
5 0

Calculating the total probability of acceptance,

𝑃 =𝑃 +𝑃

Similarly, calculate 𝑃 for other different p such as 0.01, 0.05, 0.07,

0.09, 0.11, 0.15, 0.17 etc. and plot the OC curve from the output.

Ahsanullah University of Science & Tech. Slide 20 of 29 Quality Management

10
 3/28/2021

Sequential Sampling Plans

 It is an extension of double sampling plan with a
difference in counting process.
 The acceptance or rejection number are defined as
discrete values, not fixed numbers.
 Two types of sequential sampling are available: (i)
Item by item and (ii) Group
 Item by item sampling is more common and explained
here.
 Amongst different methods, graphic scheme
illustrated by A. Wald (1947) is discussed in this slide

Ahsanullah University of Science & Tech. Slide 21 of 29 Quality Management

Sequential Sampling Plans

 It is an extension of double
sampling plan with a difference
in counting process.
 The acceptance or rejection
number are defined as discrete
values, not fixed numbers.
 Two types of sequential
sampling are available: (i) Item
by item (explained here) and (ii)
Group
 Amongst different methods,
graphic scheme illustrated by A.
Wald (1947) is discussed in this
slide
Ahsanullah University of Science & Tech. Slide 22 of 29 Quality Management

11
 3/28/2021

Sequential Sampling Plans (Cont.)

𝑋 = −ℎ + 𝑠𝑛 Acceptance Line
𝑋 = ℎ + 𝑠𝑛 Rejection Line
Where,
1−𝛼
l𝑛
𝛽 n = sequential item taken from the lot
ℎ =
𝑘
1−𝛽 h1= intercept of the acceptance limit line
l𝑛
𝛼 h2 = intercept of the rejection limit line
ℎ =
𝑘 s = slope of the limit lines
𝑝 1−𝑝
𝑘 = l𝑛 k = constant
𝑝 1−𝑝
1−𝑝
l𝑛
1−𝑝
𝑠=
𝑘

Ahsanullah University of Science & Tech. Slide 23 of 29 Quality Management

Example
 An agreement between a producer and customer states
the following data:
𝑝 = 0.01, 𝛼 = 0.05, 𝑝 = 0.1, and 𝛽 = 0.10.
Solution
𝑘 = ln . .
= ln . .
= 2.39789
.
.
𝑆𝑙𝑜𝑝𝑒, 𝑠 = = .
= 0.04
.
.
ℎ = = .
= 0.939
.
ℎ = = .
.
= 1.205

Ahsanullah University of Science & Tech. Slide 24 of 29 Quality Management

12
 3/28/2021

Example (Cont.)
So, limit lines are,
𝑋 = −ℎ + 𝑠𝑛 = −0.939 + 0.04𝑛
𝑋 = ℎ + 𝑠𝑛 = 1.205 + 0.04𝑛
Both acceptance numbers and rejection numbers must
be positive and integers. The acceptance number is the
next integer less than or equal to 𝑋 and the rejection
number is the next integer greater than or equal to 𝑋 .

Ahsanullah University of Science & Tech. Slide 25 of 29 Quality Management

Example (Cont.)
Unit Inspected Acceptance Rejection
Number Number
1 * *
2 * 2
3 * 2
4 * 2
5 * 2
6 * 2
7 * 2
10 * 2
23 * 3
24 0 3
58 0 4
83 2 5
109 3 6
Ahsanullah University of Science & Tech. Slide 26 of 29 Quality Management

13
 3/28/2021

Military and ANSI Standards

Description of the Standard
Three types of sampling are provided for:
1. Single Sampling
2. Double Sampling
3. Multiple Sampling
Provisions for each type of sampling plan include
1. Normal inspection
2. Tightened inspection
3. Reduced inspection
The quality inspection starts with Normal inspection. If
it is evident that quality of products is deteriorating,
Tightened Inspection may be followed. In case of
reverse situation, Reduced Inspection may be followed.
Ahsanullah University of Science & Tech. Slide 27 of 29 Quality Management

Switching Criteria

Description of the Standard

• Switching Rules
– Normal to tightened
– Tightened to normal
– Normal to reduced
– Reduced to normal

Ahsanullah University of Science & Tech. Slide 28 of 29 Quality Management

14
 3/28/2021

Switching Rules

Ahsanullah University of Science & Tech. Slide 29 of 29 Quality Management

15