Professional English for Software Testing
Professional English for Software Testing
This domain checks whether you can read testing literature in English: short passages, term matching and cloze items drawn from standards, tool documentation and test reports. The vocabulary is specialised rather than general, so precision matters more than reading speed.
High-frequency vocabulary clusters
- Test design: test oracle, test condition, test case, test procedure, test harness, stub, driver, traceability matrix, coverage criterion, equivalence partition, boundary value, decision table, state transition.
- Execution and defects: test log, incident report, defect vs fault vs failure (a human error introduces a fault, which may cause a failure), severity, priority, reproducibility, root cause analysis, workaround, regression, flaky test.
- Non-functional: throughput, latency, bottleneck, soak test, spike, saturation, vulnerability, exploit, mitigation, hardening, availability, mean time between failures.
- Management: entry and exit criteria, suspension criteria, lessons learned, sign-off, baseline, deliverable.
Reading strategy
- Identify the subject and main verb of long compound sentences; testing prose often buries the claim after several subordinate clauses.
- Mark negation and hedging: not necessarily, may, is unlikely to, does not imply. Options that overstate a claim are usually wrong.
- Definitional sentences (X is defined as…, X refers to…) and contrast sentences (unlike…, whereas…) are the most commonly tested.
- Distinguish error, fault and failure; verification and validation; confirmation and regression.
Worked example
Passage: Regression testing is not intended to prove that a fix works; that is the job of confirmation testing. Regression testing guards the rest of the system against unintended side effects, which is why the selection of tests should be driven by impact analysis rather than by a fixed checklist. Question: why should regression test selection be based on impact analysis? Reasoning: the passage states the purpose is to catch side effects, and it contrasts this with a fixed checklist, so the correct option is that selection must target the areas a change could affect. A distractor saying regression testing verifies the fix restates the sentence the passage explicitly denies - the most common trap in this format.
Traps
Choosing an option that is true in general but not stated in the passage; mistranslating validation as verification; treating failure, fault and defect as interchangeable; missing a double negative; ignoring the last sentence, which often carries the author's conclusion.
Key ideas
Build a personal glossary from the standards you already study in Chinese; the English and Chinese terms are usually one-to-one, which makes recall fast.