Glossary: software testing & QA A–Z

The key terms from testing and quality assurance – explained concisely and clearly.

Acceptance testing (UAT): Testing by the client or business to confirm the system meets requirements and can be accepted.
Agile testing: Testing as a continuous, whole-team activity in every sprint rather than a downstream phase. More: read article →
Black-box testing: Testing without knowledge of the internal structure – purely via inputs and expected outputs.
Boundary value analysis: A technique that tests values at the edges of ranges, where defects are especially common.
CI/CD: Continuous integration / delivery: automated building, testing and shipping on every change.
Continuous testing: Automated tests run continuously in the CI/CD pipeline, giving fast quality feedback.
Defect (Fehlerzustand): The flaw in the work product itself (e.g. in code or in a requirement). It is present from the outset and may cause a failure – but need not.
Equivalence class: A group of inputs that behave the same – one representative per class is enough and reduces the number of test cases.
Error (Fehlhandlung): The human action that produces a defect – for example a mistake in the design or a typo while coding.
Exploratory testing: Simultaneous learning, test design and execution – experience-based, without pre-scripted cases.
Failure (Fehlerwirkung): The externally visible deviation between actual and expected behaviour – what a user or a test observes. It occurs when a defect is exercised during execution.
Integration testing: Verifies the interaction of several components or systems across their interfaces.
ISTQB®: International board that publishes a vendor-neutral, modular certification scheme for testing knowledge (Foundation to Expert). More: read article →
Load and performance test: Checks behaviour, response times and stability under expected and high load.
Mock / stub: Placeholders that replace dependencies in a test to test in isolation and stably.
Regression testing: Repeated testing after changes to ensure previously working behaviour still works.
Risk-based testing: Prioritising test effort by impact and likelihood of failure. More: read article →
Smoke test: A short basic test of the main functions to check whether a build is testable at all.
Test automation: Running tests via tools instead of manually – for fast, repeatable regression. More: read article →
Test case: A concrete check with precondition, inputs, steps and expected result.
Test coverage: A measure of how much of the code, requirements or risks is covered by tests.
Test level: A level in the test process: component, integration, system and acceptance test.
Test management: Planning, steering and ownership of all test activities of a project or organisation. More: read article →
Test plan: A project-specific plan for scope, approach, resources and criteria of testing. More: read article →
Test pyramid: Distribution of tests: many fast unit tests, fewer integration, few slow UI tests.
Test strategy: The overarching guideline for how testing is done in an organisation or product. More: read article →
Unit test: A test of a single, isolated unit of code, usually automated by developers.
White-box testing: Testing with knowledge of the internal structure, e.g. to cover code paths.