Equivalence partitioning and boundary value analysis in Software Engineering - Step-by-Step Execution

Concept Flow - Equivalence partitioning and boundary value analysis

Start: Function accepts input with valid range

↓

Define input domain (e.g., age: 1-120)

↓

Equivalence Partitioning: Split into classes

↓

Invalid Low

↓

Pick ONE test from each partition

↓

Boundary Value Analysis: Test at edges

↓

Test: 0, 1, 2, 119, 120, 121

↓

Combine: partition + boundary tests

↓

End

This flow shows how equivalence partitioning divides inputs into classes and boundary value analysis tests at the edges of those classes to maximize defect detection with minimal test cases.

Execution Sample

Software Engineering

Function: validate_age(age) accepts integer 1-120

Equivalence Partitions:
  Partition 1 (Invalid Low):  age < 1    → test with -5
  Partition 2 (Valid):        1 <= age <= 120 → test with 25
  Partition 3 (Invalid High): age > 120  → test with 200

Boundary Values:
  Lower boundary: 0 (invalid), 1 (valid), 2 (valid)
  Upper boundary: 119 (valid), 120 (valid), 121 (invalid)

Final test set: {-5, 0, 1, 2, 25, 119, 120, 121, 200}

This sequence shows how to derive a minimal but effective test set using equivalence partitioning and boundary value analysis for an age validation function.

Analysis Table

Step	Action	Condition/Check	Result/Decision	Next Step
1	Define input domain	age must be integer 1-120	Domain identified	Create equivalence partitions
2	Create equivalence partitions	Split into valid and invalid classes	3 partitions: <1, 1-120, >120	Select one value per partition
3	Select partition test values	One representative per class	-5 (low), 25 (valid), 200 (high)	Identify boundary values
4	Identify boundary values	Edges of each partition	0, 1, 2, 119, 120, 121	Combine test sets
5	Combine partition + boundary tests	Remove duplicates	9 unique test values	Execute tests
6	Execute tests	Run validate_age() with each value	Verify pass/fail for each	Analyze results
7	Analyze results	Do results match expected behavior?	All correct → function validated	End

💡 9 well-chosen test values cover all equivalence classes and boundary conditions, catching most common defects without exhaustive testing.

State Tracker

Variable	Start	After Step 2	After Step 3	After Step 4	Final
partitions	None	3: <1, 1-120, >120	3 partitions	3 partitions	3 partitions
partition_tests	None	None	{-5, 25, 200}	{-5, 25, 200}	Merged into final set
boundary_tests	None	None	None	{0, 1, 2, 119, 120, 121}	Merged into final set
final_test_set	Empty	Empty	3 values	9 values	9 values tested

Key Insights - 3 Insights

Why is testing one value from each partition sufficient?

Equivalence partitioning assumes all values within a partition behave the same way. If age=25 passes validation, age=50 will too. Testing one representative per class finds the same defects as testing all values in that class.

Why do most bugs occur at boundary values?

Off-by-one errors are the most common coding mistakes. A condition like age >= 1 might be incorrectly coded as age > 1, which only fails at the boundary value 1. Boundary value analysis specifically targets these edge cases.

How many test cases does this technique save compared to exhaustive testing?

For age 1-120, exhaustive testing needs 120+ values for valid inputs alone. Equivalence partitioning plus boundary value analysis needs only 9 tests total, yet catches the same classes of defects.

Visual Quiz - 3 Questions

Look at the execution_table at step 2. How many equivalence partitions are created for validate_age()?

A 2 partitions

B 3 partitions

C 4 partitions

D 6 partitions

💡 Hint

Check the 'Result/Decision' column in step 2.

According to the variable_tracker, how many boundary test values are identified?

A 2 values

B 4 values

C 6 values

D 9 values

💡 Hint

Look at the boundary_tests row in the 'After Step 4' column.

Why is 0 included as a boundary test value but not -1?

A 0 is the value immediately below the valid lower boundary of 1

B Negative numbers cannot be tested

C 0 is always a special case in programming

D -1 is already covered by the partition test value -5

💡 Hint

Boundary value analysis tests values immediately at and around the boundary, not arbitrary values far from it.

Concept Snapshot

Equivalence partitioning splits the input domain into classes with expected similar behavior.
Boundary value analysis tests at the edges of each partition.
Combined, they produce a minimal test set that catches most common defects.
For age 1-120: 3 partitions + 6 boundary values = 9 tests instead of 120+.
Most bugs cluster at boundaries due to off-by-one errors.

Full Transcript

This visual execution shows how equivalence partitioning and boundary value analysis work together to create efficient test sets. For a function that validates age between 1 and 120, we first create three equivalence partitions: invalid low, valid, and invalid high. We pick one representative value from each partition. Then boundary value analysis adds test values at the edges: 0, 1, 2 at the lower boundary and 119, 120, 121 at the upper boundary. The combined set of 9 tests covers all input classes and boundary conditions, catching the same defects that hundreds of random tests might miss.