Key takeaways
- Levels are unit, integration, system and acceptance.
- Types describe the objective: functional or non-functional.
- The same type can be applied at different levels.
- Choose levels and types by risk, not by habit.
- Static testing such as reviews finds defects without running code.
What is levels and types of software testing?
What is the difference between levels and types?
A level answers "how big is the thing under test?". A type answers "what are we trying to learn?". For example, performance testing is a type; it can be done on a single function (unit level) or on the whole system. Mixing the two ideas is a common source of confusion in exams.
What are the four levels?
| Level | What is tested | Example |
|---|---|---|
| Unit | A function or class in isolation | A discount calculator returns the right value |
| Integration | Modules working together | Cart service calls payment service correctly |
| System | The full product in a test environment | A user completes a purchase from start to end |
| Acceptance | Readiness for business or users | Business users confirm the workflow |
| Acceptance testing includes user acceptance testing, alpha and beta testing, and contract or regulatory acceptance. |
Which integration approaches exist?
Big bang combines everything at once and makes faults hard to locate. Top-down starts with high-level modules using stubs for lower ones. Bottom-up starts with low-level modules using drivers. Sandwich mixes both. Incremental approaches make it easier to find where an interface fails.
What are functional types?
Functional testing checks behaviour against requirements. Within it you will meet smoke testing (is the build stable?), sanity testing (does a specific fix or feature work?), regression testing (did changes break existing features?), retesting (is the defect fixed?), exploratory testing (learning and testing at once) and user interface testing.
What are non-functional types?
| Type | Question |
|---|---|
| Performance | How fast is it under normal load? |
| Load | Can it handle expected users? |
| Stress | What happens beyond limits? |
| Security | Can data and access be protected? |
| Usability | Is it easy to use? |
| Compatibility | Does it work across browsers, devices, systems? |
| Accessibility | Can people with disabilities use it? |
| Reliability | Does it keep working over time? |
What is static versus dynamic testing?
Static testing examines work products such as requirements, designs and code without running them: reviews, walkthroughs and static analysis. Dynamic testing runs the software. Static testing finds defects early, when they are cheapest to fix.
How do confirmation and maintenance testing fit?
Confirmation testing, or retesting, checks that a reported defect is fixed. Regression testing checks that the fix did not break other areas. Maintenance testing happens after changes to a live system such as patches or migrations.
How do I choose what to run?
Start from risk: which failures would hurt users or the business most? Cover core flows at the system level, protect critical modules with unit and integration tests, add non-functional tests where they matter, and use acceptance tests to confirm business fit. Review the mix each release.
What mistakes appear in practice?
Testing only at the end, relying on the interface for everything, skipping integration tests, ignoring non-functional needs until late and confusing smoke with sanity. A balanced approach finds problems earlier.
How do these ideas show up in a real release?
During a sprint, developers run unit tests on every commit, the pipeline runs integration tests against a test database, testers run system tests on a staging build and the product owner performs acceptance checks. Before release the team may add performance and security checks. Seeing the same feature pass through several levels helps you explain why one kind of test is never enough, and why defects found early cost less to fix.
How do I answer level and type questions in interviews?
Define the term in one sentence, give a small example and say who usually performs it. For example, integration testing checks that modules work together, such as the cart and payment services, and developers or testers do it after unit tests. Then mention one risk it covers that other levels would miss. This three-part answer is short, clear and shows practical understanding instead of memorised definitions.
What is a short memory aid for levels?
Think of building a car: test the parts (unit), test how parts fit (integration), test the whole car on the road (system) and let the buyer test drive it (acceptance). The picture is simple and easy to recall in an exam.
What is a practical exercise for this topic?
Pick a food delivery app and list one example test for each level and for five types. Share your list with a friend and compare answers.
How do I avoid mixing up the terms?
Make flash cards with the term on one side and an example on the other, and shuffle them daily until you can answer without hesitation.
Download Software Testing Levels and Types PDF
Software Testing Levels and Types PDF
The PDF carries the same content as this page in a clean, print-ready layout, so you can read it offline and keep it for revision.
4 pages · 88 KB · PDF · Updated 5 Nov 2026