Testing versus checking is a distinction between checking, which confirms known expectations with defined outcomes, and testing, the broader work of exploring, questioning and learning about what isn't yet known.
Why it matters: Checking only covers what's already understood. Seeing the difference helps teams value exploratory work and not mistake a passing suite of checks for complete testing.
In practice: Teams automate checks for known behaviour, often as regression suites, and spend human testing time on exploration, investigation and questions that can't be scripted. Reports cover both what the checks confirmed and what exploration revealed.
Examples:
- An automated suite confirming known behaviour while a tester explores an unfamiliar area.
- A team reporting both check results and what exploration revealed.
Common pitfalls:
- Calling a green build “tested” when only checks have run.
- Using the distinction to dismiss checks, which are valuable for what they do.
Can testing be fully automated? Checking can be automated. Testing, in the wider sense of exploring and learning about what isn't known, needs human judgement, though tools can support it.
