You are on page 1of 43

Black-Box Testing Techniques II

Software Testing

Lecture 5
Cause-Effect Analysis

• Cause-Effect Analysis is a combinatorial


approach that can be viewed as a logical
extension of partition testing.
• It is a systematic means for generating
test cases to cover different
combinations of input “Causes” resulting
in output “Effects.”
Causes and Effects

• A CAUSE may be thought of as a distinct


input condition, or an “equivalence
class” of input conditions.
• An EFFECT may be thought of as a
distinct output condition or change in
program state.
Causes and Effects

• Causes and Effects are represented as


Boolean variables.
• The logical relationships among them
CAN (but need not) be represented as
one or more Boolean graphs.

Л 
Causes Effects
V 
C-E Analysis Process Steps

1. Identify Causes and Effects


– The most critical and usually the
most difficult step
– Choose an appropriate level of
abstraction.
– Divide and conquer as necessary.
– Effects may or may not be mutually
exclusive.
C-E Analysis Process Steps (cont’d)

2. Deduce Logical Relationships and


Constraints
– Relationships take the form of
conditionals and utilize the logical
operators AND, OR, and NOT.
– Constraints describe relationships
among Causes that allow for the
identification of impossible
combinations.
C-E Analysis Process Steps (cont’d)

2. Deduce Logical Relationships and


Constraints (cont’d)
– Boolean graphs provide a convenient
and economical way to visualize
relationships and constraints.
C-E Analysis Process Steps (cont’d)

3. Identify an appropriate Test Case


Selection Strategy
– Determines the number and nature
of Cause-combinations to be
considered.
– Strategies can be designed to meet a
variety of coverage requirements/
cost constraints.
C-E Analysis Process Steps (cont’d)

4. Construct a Test Case Coverage Matrix


– Typically involves tracing through
the Cause-Effect relationships to
identify combinations of Causes
resulting in each Effect according to
the selection strategy chosen.
– This can be extremely tedious...
Illustration of Step 1 (Identify Causes
and Effects)
The first input is a yes/no response to the question “Do
you reside within the city?” The second input is gross
pay for the year in question.
A non-resident will pay 1% of the gross pay in city tax.
Residents pay on the following scale:
- If gross pay is no more than $30,000, the tax is 1%.
- If gross pay is more than $30,000, but no more than
$50,000, the tax is 5%.
- If gross pay is more than $50,000, the tax is 15%.
Guidelines for identifying Causes and
Effects
• Underline words or phrases in the
specification that correspond to
input/output conditions or changes in
state.
• List each Cause and Effect.
• Assign a unique number to each (use
different number ranges to differentiate
Causes from Effects).
Guidelines for identifying Causes and
Effects (cont’d)
The first input is a yes/no response to the question “Do
you reside within the city?” The second input is gross
pay for the year in question.
A non-resident will pay 1% of the gross pay in city tax.
Residents pay on the following scale:
- If gross pay is no more than $30,000, the tax is 1%.
- If gross pay is more than $30,000, but no more than
$50,000, the tax is 5%.
- If gross pay is more than $50,000, the tax is 15%.
Illustration of Step 1 (cont’d)

Ignoring, again, the unspecified responses


to “invalid” inputs, we have:
Causes: Effects:
(1) Non-Resident (11) 1% tax
(2) Resident (12) 5% tax
(3) $0  Gross Pay  $30K (13) 15% tax
(4) $30K  Gross Pay  $50K
(5) Gross Pay  $50K
Illustration of Step 2 (Deduce Logical
Relationships and Constraints)
The first input is a yes/no response to the question “Do
you reside within the city?” The second input is gross
pay for the year in question.
A non-resident will pay 1% of the gross pay in city tax.
Residents pay on the following scale:
- If gross pay is no more than $30,000, the tax is 1%.
- If gross pay is more than $30,000, but no more than
$50,000, the tax is 5%.
- If gross pay is more than $50,000, the tax is 15%.
What are the constraints?

Causes: Effects:
(1) Non-Resident (11) 1% tax
(2) Resident (12) 5% tax
(3) $0  Gross Pay  $30K (13) 15% tax
(4) $30K  Gross Pay  $50K
(5) Gross Pay  $50K
Constraints deducible from spec,
problem domain knowledge, etc.
A. [(1) Л ¬(2)] V [¬(1) Л (2)] (i.e., one, and only one
of (1) and (2) must be true.)

B. [(3) Л ¬(4) Л ¬(5)] V [¬(3) Л (4) Л ¬(5)] V


[¬(3) Л ¬(4) Л (5)]

C. [(11) Л ¬(12) Л ¬(13)] V [¬(11) Л (12) Л


¬(13)] V [¬(11) Л ¬(12) Л (13)]
What are the logical relationships?

The first input is a yes/no response to the question “Do


you reside within the city?” The second input is gross
pay for the year in question.
A non-resident will pay 1% of the gross pay in city tax.
Residents pay on the following scale:
- If gross pay is no more than $30,000, the tax is 1%.
- If gross pay is more than $30,000, but no more than
$50,000, the tax is 5%.
- If gross pay is more than $50,000, the tax is 15%.
Conditionals deducible from
specification and constraints
• From the specification we have:

(1) => (11)


[(2) Л (3)] => (11)
[(2) Л (4)] => (12)
[(2) Л (5)] => (13)
Conditionals deducible from specification
and constraints (cont’d)
• Which, in light of the identified constraints,
simplify to:
[(1) V (3) => (11)
[(2) Л (4)] => (12)
[(2) Л (5)] => (13)
Boolean Graph Representation

Non-Res (1)

V  (11) 1% tax
[0,30K] (3)
O
(30K,50K] (4)
O
Л  (12) 5% tax
O
Res (2)
Л  (13) 15% tax
>50K (5)
Cause/Effect Constraints

Exclusive Inclusive

E “at most one” I “at least one”

One & Only One Requires

A
O “one and only one” A => B

B
Illustration of Step 3 (Identify Test Case
Selection Strategy)
• Simple (but extreme) strategies:
– “All Feasible Combinations of Cause
Values” (AFCCV)
– “All Effects covered with the
Minimum number of test Cases”
(AEMC)
• For the relationships depicted in our graph,
how many test cases would be required to
achieve AFCCV coverage? AEMC coverage?
AFCCV and AEMC Coverage

Non-Res (1)

V  (11) 1% tax
[0,30K] (3)
O
(30K,50K] (4)
O
Л  (12) 5% tax
O
Res (2)
Л  (13) 15% tax
>50K (5)
AFCCV and AEMC Coverage (cont’d)

• AFCCV:
– There are 25 = 32 possible value
combinations for all 5 Causes.
– For Causes (1) and (2), there are 2
feasible value combination pairs (due
to the “one and only one” constraint):
TF and FT.

(cont.)
AFCCV and AEMC Coverage (cont’d)

• AFCCV: (cont’d)
– Similarly, for Causes (3), (4), and (5),
there are 3 feasible value combination
triples: TFF, FTF, and FFT.
– Thus, there are 2 X 3 = 6 feasible
combinations of values for all 5
Causes, requiring a total of 6 test
cases.
AFCCV and AEMC Coverage (cont’d)

• AEMC:
– There are 3 mutually exclusive Effects.
– Thus, a different combination of Cause
values is required for each Effect to
evaluate to True.
– Therefore, 3 test cases are required.

(cont.)
AFCCV and AEMC Coverage (cont’d)

• AEMC: (cont’d)
– In general, when there are N Effects,
N or fewer test cases are required
for AEMC Coverage.
– When the N Effects are mutually
exclusive, all N test cases are
required.
AFCCV and AEMC Coverage (cont’d)

• Note that AEMC is analogous(similar) to


partitioning an input space based solely on
the specified outputs.
• AFCCV is analogous to associating a
separate equivalence class with every
(feasible) combination of the individual
input classes (i.e., the “brute-force”
approach).
• Question: do these strategies depend on
the Cause-Effect relationships?
Another Test Case Selection Strategy…

REPEAT
Select the next (initially, the first) Effect.
Tracing back through the graph (right to left), find
all feasible combinations of connected Cause
values that result in the Effect being True.
For each new such combination found:
Determine values of all other Effects, and
Enter values for each Cause and Effect in a
new column of the test case coverage matrix.
UNTIL each Effect has been selected.
What Should We Call this Strategy?

How about: All Feasible Combinations of


Connected Cause Values that Result in
Each Effect being True (AFCCCVREET)?
Applying Strategy #3

Non-Res (1)

V  (11) 1% tax
[0,30K] (3)
O
(30K,50K] (4)
O
Л  (12) 5% tax
O
Res (2)
Л  (13) 15% tax
>50K (5)
Applying Strategy #3

• Cause Value Combinations for Effect 11:


(1) V (3)  1, 3 or
1, 3 or
1, 3
• Cause Value Combinations for Effect 12:

• Cause Value Combinations for Effect 13:


Coverage Matrix
TEST CASES
CAUSES 1 2 3 4 5
Non-Resident (1) T T F
Resident (2) F F T
$0  Gross Pay  $30K (3) T F T
$30K  Gross Pay  $50 (4) F  F
Gross Pay  $50K (5) F  F
EFFECTS
1% tax (11) T T T
5% tax (12) F F F
15% tax (13) F F F
 don’t care, subject to Cause constraint B
Applying Strategy #3

Non-Res (1)

V  (11) 1% tax
[0,30K] (3)
O
(30K,50K] (4)
O
Л  (12) 5% tax
O
Res (2)
Л  (13) 15% tax
>50K (5)
Applying Strategy #3

• Cause Value Combinations for Effect 11:


(1) V (3)  1, 3 or
1, 3 or
1, 3
• Cause Value Combinations for Effect 12:
(4) Л (2)  2, 4
• Cause Value Combinations for Effect 13:
Coverage Matrix (cont’d)
TEST
CASES
CAUSES 1 2 3 4 5
Non-Resident (1) T T F F
Resident (2) F F T T
$0  Gross Pay  $30K (3) T F T F
$30K  Gross Pay  $50K (4) F  F T
Gross Pay  $50K (5) F  F F
EFFECTS
1% tax (11) T T T F
5% tax (12) F F F T
15% tax (13) F F F F
 don’t care, subject to Cause constraint B
Applying Strategy #3

Non-Res (1)

V  (11) 1% tax
[0,30K] (3)
O
(30K,50K] (4)
O
Л  (12) 5% tax
O
Res (2)
Л  (13) 15% tax
>50K (5)
Applying Strategy #3

• Cause Value Combinations for Effect 11:


(1) V (3)  1, 3 or
1, 3 or
1, 3
• Cause Value Combinations for Effect 12:
(4) Л (2)  2, 4
• Cause Value Combinations for Effect 13:
(5) Л (2)  2, 5
Coverage Matrix (cont’d)
TEST CASES
CAUSES 1 2 3 4 5
Non-Resident (1) T T F F F
Resident (2) F F T T T
$0  Gross Pay  $30K (3) T F T F F
$30K  Gross Pay  $50K (4) F  F T F
Gross Pay  $50K (5) F  F F T
EFFECTS
1% tax (11) T T T F F
5% tax (12) F F F T F
15% tax (13) F F F F T
 don’t care, subject to Cause constraint B
Complete Coverage Matrix
TEST CASES
CAUSES 1 2 3 4 5
Non-Resident (1) T T F F F
Resident (2) F F T T T
$0  Gross Pay  $30K (3) T F T F F
$30K  Gross Pay  $50K (4) F  F T F
Gross Pay  $50K (5) F  F F T
EFFECTS
1% tax (11) T T T F F
5% tax (12) F F F T F
15% tax (13) F F F F T
 don’t care, subject to Cause constraint B
Comparing Strategy #3 to AFCCV and
AEMC
• How does Strategy #3 differ from AFCCV?
– For each Effect, only the connected
Causes are considered.
– It is less conservative:
• Does not ensure that every feasible
combination of Cause values will be
covered.
• And thus does not ensure that every
feasible combination of Effect values will
be covered. (Relevant when Effects are
not mutually exclusive.)
Comparing Strategy #3 to AFCCV and
AEMC
• How does it differ from AEMC?
– It is more conservative (ALL feasible
combinations of connected Cause
values must be covered for each
Effect).
Coming up in “Black-Box Testing
Techniques III”…
• We step-through another (somewhat
more complex) example of Cause-Effect
Analysis.
• Describe a test case design technique for
exploring boundary conditions.
• Consider a test case design strategy
based on intuition and experience.

You might also like