Interview questions / Foundations
Levels, types and techniques of testing: the vocabulary interviewers expect you to use precisely. Expect these in screening rounds for every QA and SDET role.
6 questions · 3 beginner· 1 intermediate· 2 advanced
beginner 0–2 yrs
Verification asks 'are we building the product right?': does the work product meet its specification, checked through reviews, static analysis and many tests. Validation asks 'are we building the right product?': does it meet the user's actual needs, checked through UAT, demos, beta and usability testing.
Full answer with examplesbeginner 0–2 yrs
Smoke testing is a shallow, broad check that a build is stable enough to test at all. Sanity testing is a narrow, focused check that a specific fix or change works. Regression testing re-checks existing functionality to make sure recent changes did not break it, and is the main target for automation.
Full answer with examplesEquivalence partitioning groups inputs the system should treat the same, so one value per partition is enough. Boundary value analysis tests the edges of each partition, where off-by-one defects cluster. For an age field accepting 18 to 60, test 17, 18, 60, 61 plus valid and invalid representatives.
Full answer with examplesintermediate 2–5 yrs
The test pyramid recommends many fast, isolated unit tests, fewer integration or service tests, and a small number of slow end-to-end UI tests, to get fast, reliable feedback cheaply. The principle still holds, but the shape is a heuristic: the testing trophy and honeycomb shift weight to integration tests.
Full answer with examplesAutomate checks that are repeated often, deterministic, stable in requirements, high-risk, data-heavy or tedious, at the lowest layer that catches the defect. Keep humans on exploratory, usability, one-off and fast-changing areas. Weigh build and maintenance cost against how often the test gives useful feedback.
Full answer with examplesCoverage shows what code was executed, not what was verified, so it finds untested areas but cannot prove tests are good. Better signals combine mutation score, escaped defects, flakiness rate, feedback time and coverage of risks and requirements. Use metrics to ask questions, not as targets.
Full answer with examplesScreening rounds for QA and SDET roles almost always include software testing interview questions on fundamentals. They are quick to ask and reveal a lot: candidates who use terms loosely (calling every quick check "smoke testing") or recite definitions without examples tend to struggle later in the loop. The questions here cover the vocabulary and techniques interviewers expect you to use precisely, with answers that give an example for each definition.
Two questions in this set are really about judgement. "Is code coverage a good measure of test quality?" expects you to say that coverage shows what was executed, not what was checked: a test with no assertions can reach 100% line coverage. Better signals include mutation testing, escaped defects and how often failures point to real bugs. "Which tests should you automate?" expects a decision framework: stable, repeatable, high-value checks that run often are good candidates; one-off exploratory work and rapidly changing UI usually are not.
Give the definition in one sentence, follow it with a concrete example from an application the interviewer will recognise, and finish with when it matters in practice. For "what is regression testing?" that might be: re-running existing tests after a change to confirm nothing that worked before is broken; for example, re-checking checkout after a change to the discount service; and in practice the regression suite is the main candidate for automation because it runs on every release.
Avoid reciting textbook lists such as "the seven principles of testing" unless asked. Interviewers remember the candidate who explained one principle with a real example, not the one who listed all seven.
Work through the Software Testing Fundamentals learning path if any term is unfamiliar, then answer these questions without notes. The boundary value calculator and pairwise generator let you apply the design techniques to your own inputs, and the manual testing interview questions put the same concepts into scenarios.
