Structured interview scorecards that preserve evidence

A structured interview scorecard defines the dimensions that matter for a role and the evidence reviewers should use to assess them. It creates a common language before the first candidate is discussed.

The scorecard is not the interview transcript and it is not a spreadsheet of arbitrary numbers. It is a compact decision record: criterion, evidence, interpretation, confidence, and unresolved questions.

Begin with the role outcomes, then select a small set of dimensions broad enough to hold meaningful evidence and distinct enough to avoid double counting.

A balanced set might include

• Role fit: relevant experience, craft judgment, and domain requirements.
• Problem solving: framing, trade-offs, iteration, and learning.
• Ownership: responsibility, follow-through, and response to setbacks.
• Communication: clarity appropriate to the work and audience.

RoundZero reports use these broad dimensions alongside role-specific evidence. Your own scorecard may need a technical, operational, leadership, safety, or customer dimension when it is materially different from the four above.

Labels such as “excellent” or “poor” invite each reviewer to use a different standard. Describe what a reviewer would observe instead.

• Strong evidence: a specific example, clear individual contribution, sound trade-offs, and an outcome connected to the criterion.
• Mixed evidence: relevant exposure with unclear ownership, shallow reasoning, or an outcome that cannot be separated from the team.
• Concerning evidence: repeated vagueness, material contradiction, unsafe reasoning, or no example after reasonable follow-up.

Add “not enough evidence” as a valid state. It is more honest than forcing uncertainty into the middle of a numeric scale.

Every core question should cover at least one dimension, and every essential dimension should have a planned way to collect evidence. Avoid a long bank of questions with no relationship to the final review.

1. Ask for a specific situation tied to the work.
2. Clarify the candidate’s personal responsibility and constraints.
3. Probe one consequential decision or trade-off.
4. Ask what changed, what the result was, or what they would do differently.

Comparable coverage matters more than identical wording. Adaptive follow-up can clarify evidence while the scorecard keeps the underlying criteria stable.

Link each score or label to the answer, excerpt, work sample, or observed behavior that supports it. Separate a candidate’s claim from the reviewer’s interpretation.

RoundZero candidate reports follow this pattern with a summary, strengths, weaknesses, insights, supporting excerpts, screening-question review, authenticity signals, and dimension-level scores. The voice step adds separate observations for clarity, articulation, conciseness, listening, and confidence; it does not replace role evidence.

Ask each reviewer to complete their evidence notes before the debrief. Start the meeting with criteria and exceptions, not with the loudest overall recommendation.

• Discuss dealbreakers and missing evidence explicitly.
• Distinguish a job requirement from a coachable gap.
• Record why the final decision differs from an initial recommendation.
• Use rankings to organize review, not to pretend close scores are exact.

1. Dimension: the job-relevant capability being assessed.
2. Evidence expected: observable strong, mixed, and concerning signals.
3. Question coverage: core question and permitted follow-up areas.
4. Evidence captured: excerpt or concise factual note.
5. Assessment: score or label, confidence, and rationale.
6. Open question: what still needs human follow-up.

• Structured Interview Guide — U.S. Office of Personnel Management
• Structured Interviews: Assessment Policy FAQ — U.S. Office of Personnel Management
• Uniform Guidelines on Employee Selection Procedures — Electronic Code of Federal Regulations