Glossary (Quick, Practical)

Short, practitioner-friendly definitions you can use during design reviews and when writing cases. Tip: skim this once; link back when terms pop up in playbooks or checklists.

Notation & Abbreviations

MAE — Main / Alternative / Exception (scenario families).
MECE — Mutually Exclusive, Collectively Exhaustive (no overlap, no gaps).
SUT — System Under Test.
I/O — Input / Output.
RACI — Responsible, Accountable, Consulted, Informed (ownership).

Core Test-Design Concepts

Coverage thinking — Structuring the space of behaviors so tests are representative and sufficient (not every permutation).
Equivalence classes — Group inputs that should behave the same; test one per group + boundaries.
Boundary value analysis — Hit “just in / just out” values around each limit (e.g., −1, 0, 1).
Decision table — Tabular rules: conditions × actions. Ensures no conflicting or missing rules.
State model — Valid states + transitions + guards; tests verify allowed/blocked moves and side-effects.
Pairwise testing — Cover all pairs across factors; reduces explosion while catching many interaction bugs.
CRUD grid — For each resource: Create/Read/Update/Delete × roles × constraints.
Oracle (test oracle) — Mechanism to decide pass/fail (calc, invariant, golden file, service contract).
Traceability — Links across Requirement ↔ Scenario ↔ Case ↔ Evidence; often a matrix (RTM).
Risk-based prioritization — Test higher business/technical risk first (impact × likelihood).

Scenario & Workflow

Main/Alt/Exception flows (MAE) — Happy path, common variations, and error/edge paths.
Invariants — Conditions that must always hold (e.g., sum of line items = order total).
Idempotent action — Safe to repeat: same key/request ⇒ same outcome (e.g., retry POST with idempotency key).
Race condition — Outcome depends on timing/order of concurrent events.
Idempotent consumer — Consumer deduplicates replays, guaranteeing single-effect processing.

API & Data Contracts

Contract test — Verify provider/consumer agree on schema, types, required fields, and error shapes.
Schema validation — Enforce structure (e.g., JSON Schema/OpenAPI) before deeper checks.
Pagination (offset vs cursor) — Offset is page index; cursor is opaque pointer for stable traversal.
Consistency guarantees — Strong / Read-after-write / Eventual; dictates assertion timing.
Idempotency key — Client-generated token to dedupe retried operations.
Error taxonomy — Canonical set of errors (auth, validation, rate-limit, transient) mapped to UX messages.
HTTP semantics — Safe (GET), idempotent (PUT/DELETE), unsafe (POST) per RFC intent.
ETag / If-Match / If-None-Match — Conditional requests for concurrency control & caching.
Pagination invariants — Sorting stability or cursor monotonicity to avoid skips/dupes across pages.

p95 / p99 — 95th/99th percentile latency; budgets set expectations beyond average.
TTFB / TTI — Time to First Byte / Time to Interactive (web UX timing signals).
SLO / SLI / SLA — Objective / Indicator / Agreement for reliability commitments.
Budget — Agreed threshold (e.g., p95 ≤ 500 ms) used as an acceptance criterion.
Retry w/ backoff — Reattempt with increasing delay; add jitter to avoid thundering herds.
Circuit breaker — Stop calling a failing dependency until it recovers to protect the system.
Graceful degradation — Reduced functionality with clear UX when dependencies fail.
Compatibility matrix — Supported OS/Browser/Device/App versions and features.
Internationalization (i18n) / Localization (l10n) — Build for multiple locales / adapt strings, formats, rules.

Observability & Evidence

Structured logs — Key-value logs parsable by machines (include correlation/trace IDs).
Metrics — Numeric time-series (counters, gauges, histograms) for SLOs and budgets.
Tracing — End-to-end spans for a request; proves where time/errors occur.
Correlation ID — ID that ties logs/metrics/traces for a single workflow.
Evidence artifact — Saved proof: log snippet, screenshot, export, API transcript validating the expected result.

Security & Privacy

Authentication vs Authorization — Who you are vs what you may do.
Least privilege — Grant minimal permissions needed for a task.
Input validation & encoding — Validate on trust boundary; encode at sink (HTML/SQL/OS).
Secret handling — No secrets in logs; use vaults/kms; rotate regularly.
PII / Data minimization — Store the least personal data necessary; mask in lower environments.
Audit trail — Tamper-evident record of who did what, when, and why.

Accessibility (WCAG)

POUR — Perceivable, Operable, Understandable, Robust (WCAG principles).
Keyboard accessibility — Full operation without a mouse (focus order, visible focus).
ARIA semantics — Roles/states to aid assistive tech (use sparingly and correctly).
Color contrast — Sufficient luminance difference for text/UI elements.
Alt text & labels — Programmatic names for images, inputs, controls.

AI-Augmented Test Design (Optional)

Grounding (explicit-only) — Use only provided requirements/artifacts as sources; no outside guessing.
Schema discipline — Outputs adhere to a specified JSON/CSV/Markdown schema.
Deduplication — Avoid semantic or structural duplicates in generated scenarios/cases.
Human-in-the-loop — Reviewer notes → regenerate loop to refine outputs.
Hallucination — Fabricated content not present in sources; must be prevented by guardrails.

Review Gates & Quality Signals

Gate — A checklist or automated criterion that must pass before merging/releasing.
Defect leakage — Escaped defects per release; track ↓ as design quality ↑.
Case effectiveness — Share of defects caught by designed tests vs ad-hoc discovery.
Coverage diff — Change in scenario/case coverage between versions (what’s new or removed).

Tiny Examples (copy-pasteable)

Boundary set (inclusive 0–100): -1, 0, 1, 99, 100, 101
Idempotency test: Send POST /payments with Idempotency-Key: K1 twice → same result; with K2 → new payment.
Retry/backoff: 100ms, 200ms, 400ms (+ jitter); abort after N attempts or on non-retryable errors.
Error taxonomy → UX: VALIDATION_ERROR.amount.invalid ⇒ “Enter a whole number between 0 and 10,000.”

Glossary

Glossary (Quick, Practical)

Notation & Abbreviations

Core Test-Design Concepts

Scenario & Workflow

API & Data Contracts

Observability & Evidence

Security & Privacy

Accessibility (WCAG)

AI-Augmented Test Design (Optional)

Review Gates & Quality Signals

Tiny Examples (copy-pasteable)

See Also

On this page