Verification and Validation

Ensuring software quality requires rigorous verification and validation (V&V) activities. These techniques help detect defects, measure quality, and build confidence that software meets its requirements and user needs.

V&V: What’s the Difference?

Verification

“Are we building the product right?”

Verification checks whether software conforms to its specifications:

Does the implementation match the design?
Does the design satisfy the requirements?
Are we following standards and processes correctly?

Validation

“Are we building the right product?”

Validation checks whether software meets user needs and expectations:

Does the software solve the actual problem?
Will users be satisfied with the solution?
Does it provide value in the real world?

Why V&V Matters

Software defects are expensive and dangerous:

Cost escalation: Defects found in production cost 100x more to fix than those caught early
Safety risks: In critical systems (medical, automotive, aerospace), defects can be fatal
Business impact: Quality issues damage reputation, lose customers, and reduce revenue
Maintenance burden: Poor quality code becomes harder and more expensive to maintain

Prevention is cheaper than cure - investing in V&V activities early saves time and money later.

Topics in This Section

V&V Overview

Understand the landscape of verification and validation techniques:

Techniques - Static vs dynamic, manual vs automated approaches
Testing - Black box, white box, levels, and purposes
Static Analysis - Code reviews, inspections, and automated analysis tools
Inspection - Formal review processes

Test Coverage Criteria

Learn how to measure and achieve thorough testing:

Terminology - Key concepts in coverage analysis
Code Coverage - Statement, branch, and path coverage
Condition Coverage - Decision, condition, and MC/DC coverage
Basis Path - Cyclomatic complexity and independent paths
Data Flow - Definition-Use pairs and variable lifecycle
Mutation Testing - Measuring test suite effectiveness
Discussion - Trade-offs and practical considerations

Random Testing

Techniques that generate test inputs by sampling rather than systematic design:

Fundamentals - Theory, reliability prediction, Randoop, AutoTest
Fuzz Testing - From Miller 1990 to coverage-guided greybox fuzzing
Property-Based Testing - QuickCheck, Hypothesis, JQF, industry practice
Mutation Testing - Fault seeding, equivalent mutants, LLM-assisted detection

Exploratory Testing

Human-driven testing where learning, test design, and execution happen simultaneously:

Techniques and Heuristics - Whittaker’s tours, Hendrickson’s heuristics, dynamic test design
Session-Based Test Management - Charters, time boxes, metrics, degrees of exploration
Empirical Evidence - ET vs. scripted experiments, efficiency data, tester knowledge

Operational Profile

Usage-based testing proportional to how software is used in the field:

Developing an OP - Musa’s five-step procedure, representations, data sources
Industrial Evidence - AT&T, HP results, Android profile coverage
Modern Approaches - Operational coverage, adaptive testing, relative coverage

Testing Approaches

Black Box Testing (Functional)

Testing based on requirements and specifications without knowledge of internal structure:

Focus on inputs, outputs, and behavior
Tests what the software does, not how it does it
Examples: Equivalence partitioning, boundary value analysis, use case testing

White Box Testing (Structural)

Testing based on internal code structure and logic:

Focus on paths, conditions, and control flow
Tests how the software implements functionality
Examples: Statement coverage, branch coverage, path testing

Static Analysis

Examining code without executing it:

Code reviews and inspections
Automated tools (linters, type checkers, static analyzers)
Detect issues early, before runtime

Testing Levels

Software testing occurs at multiple levels:

Unit Testing - Testing individual components in isolation
Integration Testing - Testing interactions between components
System Testing - Testing the complete integrated system
Acceptance Testing - Validating software meets user needs

Key Principles

Testing shows presence of defects, not absence - You can’t test quality in; you must build it in
Early testing saves time and money - Shift-left: test as early as possible
100% coverage is impossible - Focus on risk-based testing
Pesticide paradox - Running the same tests repeatedly finds fewer new bugs
Context matters - Testing approach depends on criticality, domain, and constraints
Automate when possible - Automated tests enable continuous quality feedback

Test Adequacy

How do you know when you’ve tested enough?

Coverage criteria provide objective measures:

Structural coverage (statements, branches, paths covered)
Functional coverage (requirements, features tested)
Mutation score (test suite effectiveness)

But adequacy also depends on:

Risk - Critical functionality needs more testing
Cost - Balance testing investment with value
Time - Testing under deadlines requires prioritization

Further Exploration

After mastering V&V techniques, explore:

Defining Quality - Understanding what quality means
Organizing Quality - Managing quality processes
Quality Attributes - Achieving specific quality goals

Disclaimer: AI is used for text polishing and explaining. Authors have verified all facts and claims. In case of an error, feel free to file an issue.