How to Create an Interview Scoring Rubric That Actually Works

A scoring rubric without anchors is just a number scale. It doesn't tell two different interviewers whether a candidate's answer deserves a 2 or a 3 — and that ambiguity is where bias creeps in. Here's how to build rubrics that actually standardise your interviews.

Why Most Scoring Systems Fail

Interviewers given a 1–5 scale with no guidance default to their gut. One interviewer's "3" is another's "4". Candidates who interview later in the day get lower scores because interviewers are fatigued (contrast effect). Candidates who remind you of yourself get the benefit of the doubt on borderline answers (affinity bias).

Behavioural anchors fix this by describing — in words — what each score point looks like for each specific question.

The 1–4 Scale: Why Four Points

A 4-point scale forces a decision. On a 5-point scale, interviewers cluster around the middle (3). A 4-point scale eliminates the neutral option. Either the answer is below standard (1–2) or above (3–4). This creates more differentiation between candidates.

Writing Behavioural Anchors

For each question, write a description of what a 1, 2, 3, and 4 answer actually looks like. The anchors should be:

• Specific to the question — not generic praise or criticism
• Observable — based on what the candidate says, not whether you liked them
• Graduated — each score point should be clearly distinct from the others

Example: "Tell me about a time you dealt with an underperforming team member."

• 1 — Poor: Unable to give a specific example, or describes avoiding the conversation entirely.
• 2 — Below standard: Gives a vague example with no clear process; outcome unclear or negative.
• 3 — Good: Describes a specific situation, demonstrates direct but fair conversation, can explain outcome.
• 4 — Excellent: Clear process (documented, progressive), shows empathy and commercial awareness, outcome demonstrates measurable improvement or well-managed exit.

Deal-Breakers vs Low Scores

Some answers aren't just weak — they're disqualifying. Distinguish between a 1 (poor but not a dealbreaker) and a genuine dealbreaker that ends the process. Examples of dealbreakers: admitting to unsafe behaviour, demonstrating no awareness of direct reports as individuals, inability to give any management example at all.

Calibrating Your Panel

Even excellent anchors need calibration. Before your first interview, have your panel independently score two sample answers (written or recorded), then compare. Discuss any discrepancies. This calibration exercise reduces inter-rater variance dramatically and usually takes under 20 minutes.