Debug precision and missed coverage
Course overview · 4 min reading + 12 min practice, estimated
Principles and method
Inspect a consistent sample of results and label relevant, irrelevant and uncertain cases using the brief. Irrelevant results reveal overly broad concepts or ambiguous terms. Missing known relevant examples reveal overly restrictive conditions or weak terminology. Precision describes the proportion of retrieved results that are relevant under your review rule. Recall concerns the relevant population retrieved, but usually cannot be calculated exactly without a known reference set. Avoid claiming complete coverage from a search interface. Change one major query element at a time so the effect can be interpreted.
Worked example
A query for QA returns software testers and manufacturing quality staff. Adding production context improves relevance, but requiring one exact title removes a known suitable quality manager. The search log records both effects.
Put it into practice
Review ten fictional results and calculate precision, then identify one query change.
Use fictional information and keep your work in your own notes.
Compare your approach: self-review guidance
If six of ten reviewed results meet the rule, sample precision is 60%. Explain that this says nothing by itself about how many relevant people were missed. Include uncertain cases transparently rather than silently counting them as successes.
Sources and further reading
Original Academy teaching and fictional examples. These references provide context, not endorsement. Edition 2026.09; updated 2026-09-24.
- CIPD: Recruitment
Professional context for the recruitment lifecycle.
- ICO: Recruitment and selection
UK guidance. Check its current status and updates before implementation.