| name | structured-interview-kit |
| description | Use when designing or revising a role's interview loop — builds a competency-to-stage map, standardized questions, leveled rubrics, and interviewer instructions so evaluation is consistent and bias-resistant. |
Structured Interview Kit
A playbook for building an interview loop that measures the same thing, the same way, for every candidate. The failure mode it prevents: five interviewers asking overlapping questions, grading on gut feel, and a debrief that converges on whoever spoke first.
Map competencies to stages
- List the must-have competencies from the scorecard. These, and only these, are what the loop grades.
- Assign each competency to exactly one stage and one interviewer. Full coverage, zero overlap — if two stages assess the same competency, merge or reassign until each is owned once.
- Confirm every competency has an owner and no interviewer is asked to grade more than they can observe well in one session.
Write questions and rubrics
- For each competency, write a primary question and two or three follow-ups. Prefer "walk me through a time you actually…" over hypotheticals; past behavior and real work samples predict better than speculation.
- Write a leveled rubric per competency — strong / adequate / weak — with a concrete description of what an answer at each level contains. The test: two interviewers grading the same transcript should land on the same level.
- Standardize the questions across candidates for a role. The candidate varies; the questions don't.
- Where the competency is a skill, design a time-bounded work sample with its own grading guide, so ability shows through polish.
Remove bias by construction
- Strike prompts that reward pedigree, culture-sameness, or quick talkers over the actual competency.
- Instruct interviewers to record observations separately from judgments, and to grade independently before any debrief discussion.
- Include a short interviewer brief: their competency, their questions, their rubric, and what earlier stages already covered so they don't re-litigate it.
Quality bar
- Every scorecard competency is covered exactly once.
- Rubrics are concrete enough that two strangers agree on a grade.
- Questions are standardized and free of leading or pedigree cues.
- Interviewers grade independently first — the kit enforces it.
Output format
Interview kit: competency-to-stage map → per-stage question sets with leveled rubrics → work-sample exercises with grading guides → interviewer briefs and independent-grading instructions.
