Blog
EducationLearnLift AIOctober 28, 20266 min read

Common QA Testing Terms Every Beginner Should Know

A beginner-friendly glossary of QA testing terms, from test cases and bugs to regression, smoke testing, and acceptance criteria.

If you are starting in software testing, the vocabulary can feel harder than the first test itself. Teams use short phrases such as “run a smoke test,” “cover the edge cases,” or “verify the acceptance criteria” as if everyone already knows what they mean.

You do not need to memorize a dictionary before you can learn QA. You need a working set of terms that helps you understand a ticket, describe a defect, and explain your reasoning in an interview. This glossary covers the concepts beginners meet most often, with small examples you can reuse while practicing.

The basic language of testing

Quality assurance (QA)

Quality assurance is the broader process of improving confidence in a product. It includes planning, reviewing requirements, designing tests, reporting findings, and checking whether the team’s process catches problems early. QA is not only “clicking through the app” at the end of development.

Quality control (QC)

Quality control focuses on checking the product itself. Executing a test case and recording whether a login form behaves correctly are QC activities. In everyday conversation, people often use QA and testing interchangeably, but QA describes the wider quality process.

Test scenario

A test scenario is a high-level situation you want to examine. For an online shop, “customer checks out with a saved card” is a scenario. It gives direction without listing every click.

Test case

A test case is a repeatable set of steps, data, and expected results for checking one behavior. A login test case might specify a valid email, a valid password, the action of selecting Sign in, and the expected dashboard. Good test cases make the result understandable to another tester.

Test data

Test data is the information used during a test: an account, address, date, product, or deliberately invalid value. Include the important properties of the data so someone else can reproduce your result. “A long password with spaces” is more useful than “bad password.”

Expected result and actual result

The expected result is what the requirement says should happen. The actual result is what the software did. Comparing the two is the heart of a test. If the expected result is “an error message appears and the form remains open” but the page silently refreshes, the difference is evidence worth recording.

Terms for defects and requirements

Bug, defect, and failure

A bug or defect is a flaw in the product or its implementation. A failure is the visible incorrect behavior when that flaw is triggered. Many teams use the words casually as synonyms. In a report, focus on the observable behavior and the steps that produced it rather than debating the label.

Reproduction steps

Reproduction steps are the shortest reliable path to seeing a problem again. They should include setup, actions, and relevant data. A precise sequence such as “open Settings, change language to Spanish, close and reopen the app” helps a developer investigate much faster than “language is broken.”

Severity and priority

Severity describes the impact of a defect on the system or user. Priority describes how soon the team chooses to address it. A typo on a public checkout button may have low technical severity but high priority because it affects many customers. Keep the two ideas separate when discussing a bug.

Acceptance criteria

Acceptance criteria are the conditions a feature must meet to be considered complete. They often appear in a user story. For a password reset feature, criteria might say that a valid email receives a link, an unknown email receives a safe response, and an expired link cannot set a password. Acceptance criteria become useful test ideas.

Edge case

An edge case is an unusual value or situation near the boundary of expected use. Examples include an empty required field, a file at the maximum allowed size, February 29, or a user who has just lost network access. Edge cases are valuable because assumptions tend to fail at boundaries.

Common test types

Functional testing

Functional testing checks what a feature does. Does a search return matching products? Does saving a profile preserve the new phone number? The test is based on behavior and requirements, not on the internal code structure.

Non-functional testing

Non-functional testing examines qualities such as performance, accessibility, security, compatibility, and usability. A page that returns the correct data but takes 20 seconds to load may pass a narrow functional test and still fail an important user need.

Smoke testing

Smoke testing is a short, broad check that a build is stable enough for deeper testing. You might open the app, sign in, navigate to a key page, and complete one critical action. If these basics fail, spending hours on detailed scenarios is usually premature.

Sanity testing

Sanity testing is a focused check of a particular change or fix. After a developer corrects a date filter, you may verify that filter and a few nearby behaviors. The exact distinction from smoke testing varies by team, so learn how your project uses the terms.

Regression testing

Regression testing checks that existing behavior still works after a change. A new checkout discount could affect totals, refunds, order history, and receipts. Regression tests help uncover those unintended side effects.

Retesting

Retesting means running the failed test again after a fix. First confirm the original defect no longer occurs. Then run relevant regression checks to see whether the fix introduced another problem. Retesting and regression testing answer different questions.

Exploratory testing

Exploratory testing combines learning, test design, and execution. Instead of following only prewritten steps, you investigate the product, form a hypothesis, and follow useful clues. Take notes about what you tried so a promising discovery can become a repeatable test case.

A simple way to study the vocabulary

Start with one feature, such as login, and write a small map:

  • scenario: a user signs in;
  • test case: valid credentials open the dashboard;
  • test data: a verified account and an incorrect password;
  • expected result: a clear message appears for the incorrect password;
  • edge case: the email field contains spaces;
  • regression risk: password reset or session handling changes.

Then explain the map aloud in your own words. If you can connect a term to an action and an expected result, you are learning a skill rather than reciting a definition. Keep a short list of terms you confuse and review it after a practice session.

Where a study coach fits

Short, repeated recall is often easier to maintain than one long glossary session. LearnLift AI can fit this workflow by turning QA concepts into review prompts, flashcards, and quizzes, while leaving you to check whether you can apply each term to a realistic scenario. Use it to identify weak areas, then return to a test case or bug report and practice the idea in context.

The goal is not to sound impressive by using every acronym. It is to communicate clearly: describe the risk, show the steps, compare expected with actual behavior, and explain what you would check next. That foundation will make both your interview answers and your day-to-day QA work more precise.

Related posts

Keep reading.

All posts
Loading page…