Thanks to visit codestin.com
Credit goes to www.scribd.com

Scribd
Upload
171 views26 pages

Pairwise Testing Techniques Guide

Pairwise testing is a technique that can significantly reduce the number of test cases needed compared to traditional exhaustive testing. It involves testing all possible pairs of each parameter rather than all possible combinations. There are two main techniques for generating pairwise test cases: orthogonal arrays and all-pairs algorithms. Orthogonal arrays arrange test parameters into a table to efficiently generate test cases covering all pairs. All-pairs algorithms use programs to directly generate the pair combinations without arrays. While pairwise testing covers most bugs, it does not guarantee finding all failures and some generated test combinations may not be realistic.

Uploaded by

vineeta1234
Copyright
© Attribution Non-Commercial (BY-NC)
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PPTX, PDF, TXT or read online on Scribd
171 views26 pages

Pairwise Testing Techniques Guide

Pairwise testing is a technique that can significantly reduce the number of test cases needed compared to traditional exhaustive testing. It involves testing all possible pairs of each parameter rather than all possible combinations. There are two main techniques for generating pairwise test cases: orthogonal arrays and all-pairs algorithms. Orthogonal arrays arrange test parameters into a table to efficiently generate test cases covering all pairs. All-pairs algorithms use programs to directly generate the pair combinations without arrays. While pairwise testing covers most bugs, it does not guarantee finding all failures and some generated test combinations may not be realistic.

Uploaded by

vineeta1234
Copyright
© Attribution Non-Commercial (BY-NC)
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PPTX, PDF, TXT or read online on Scribd
You are on page 1/ 26

Pairwise testing

(a powerful technique)
Key points
Understanding what the challenges are with test
design
Master the 2 pairwise design techniques; Orthogonal
Arrays and All-pairs algorithm
Learn about the powers and dangers of this technique
Understanding Challenges
 A simple mobile phone
 3 call types;
free, chargeable (local) and chargeable (international)
 3 message types;
text, mms and video
 2 account types;
pay as you go and contract
Understanding Challenges
Car Insurance
 3 policy types (TP, TPFT, FC)
 3 storage modes (garaged, drive, road)
 4 NCD (0, 1, 2 and 3+ years)
 2 licence types (full and provisional)
 5 age categories (17-21, 22-30, 31-40, 41-50 and 51+)
 5 engine sizes (<1000, 1000-1600, 1601-2000, 2001-
2999 and 3000+)
Understanding Challenges

So how do we select a subset of test cases?


The good news ☺

4 parameters each have 3 values


= 3 x 3 x 3 x 3 = 81 test cases
pairwise = 9 test cases
13 parameters each have 3 values
= 3 x 3 x 3 x 3…= 1,594,323 test cases
pairwise = 15 test cases
 20 parameters each have 10 values
= 10 x 10 x 10 x 10….= a lot of test cases !!!
pairwise = 180 test cases
Why does pairwise testing work?
The majority of faults are either single or double
mode

There are 2 recognised methods for generating all the


pairs
Orthogonal arrays
All pairs algorithm
Orthogonal Arrays
Orthogonal Arrays
Example 1
 A simple mobile phone
3 call types;
free, chargeable (local) and chargeable (international)
 3 message types;
text, mms and video
 2 account types;
pay as you go and contract
Using Orthogonal Arrays
Step 1: identify the variables
call type, messaging and account type = 3
Step 2: determine the number of choices for
each variable:-
call type: free, local, international = 3
messaging: text, mms, video = 3
account type: pay as you go, contract = 2
Contd..
Step 3: locate the orthogonal array which has a
column for each variable and values within the
columns that correspond to the choices for each
variable
Contd..
Contd..
Step 4: map the problem onto the array

var 1 = call type 1=free, 2=local, 3=international


var 2 = messaging 1=text, 2=mms, 3=video
var 3 = account type 1=PAYG, 2=contract, 3=don’t care
Contd..
Contd..
Step 5: construct the test cases
Example 2
Car Insurance
3 policy types (TP, TPFT, FC)
3 storage modes (garaged, drive, road)
4 NCD (0, 1, 2 and 3+ years)
2 licence types (full and provisional)
5 age categories (17-21, 22-30, 31-40, 41-50 and 51+)
5 engine sizes (<1000, 1000-1600, 1601-2000, 2001-2999
and 3000+)
example 2
ALL PAIR ALGORITHIM
Simplified way of generating all pairs combinations
Does not need external devices such as orthogonal
arrays
Usually produced using a program
http://www.satisfice.com
http://www.pairwise.org/about.asp
Mobile phone example using allpairs
So what’s the difference?
In some instances they produce the same number of
test cases
In some instances the all-pairs algorithm
produces fewer test cases
In some instances orthogonal arrays produce
fewer test cases
What experts say
Some Statistics
Running 1% to 20% of all possible tests you will
find 70% to 85% of the total bugs
Cohen reported that test cases created by the “allpairs”
algorithm provided better code coverage than random
tests
 300 random tests: 67% statement coverage and
58% decision coverage
200 “all-pairs” tests: 92% statement coverage and
85% decision coverage
A word of caution
Do not test all combinations
May not produce the most common combinations
No priority is given to the tests
Some combinations produced may be infeasible
Q&A
THANK YOU!

You might also like