Methodology that makes interviews comparable.
Proofglint's structured interview module anchors every question to observable behavior examples — so scores mean the same thing to every interviewer, every time.
BARS: the gold standard of structured interview design.
Behaviorally Anchored Rating Scales (BARS) attach specific, observable behavior examples to each score point on a rating scale. Instead of asking "was that answer good or bad?" interviewers match what they heard to pre-written anchors.
The result: scores are calibrated before the interview starts. Two interviewers watching the same answer will converge on the same score — not because they agreed in the room, but because the rubric did the calibration work in advance.
Read the full Rubric GuideFrom rubric to decision in four steps.
Build Your Rubric
Create scoring dimensions for the role. Write behavioral anchors for each score level using BARS methodology.
Assign Interviewers
Assign interviewers to each loop stage. They see the questions and anchors — but not each other's scores.
Conduct & Score
Each interviewer submits scores and evidence notes after their session. Scores are locked until debrief.
Debrief Calibration
All scores revealed side by side. Outliers flagged. Hiring decision logged with rationale — immutably.
Everything structured interviews need. Nothing they don't.
BARS Rubric Builder
Define scoring dimensions and write behavioral anchors for each 1–5 level. Reuse rubrics across roles.
Multi-Interviewer Assignment
Assign different interviewers to different stages. Each scores independently, preventing anchoring bias.
Blind Pre-Debrief Scoring
Scores are hidden from all interviewers until the debrief begins. Prevents groupthink and anchoring effects.
Evidence Notes
Each score requires a brief evidence note. What did the candidate say that led to this score? Documented.
Loop Debrief Aggregation
All scores, averages, and outlier flags shown in a single calibration view. The debrief actually runs itself.
Immutable Decision Record
Final hiring decision and rationale are logged and locked. EEOC-relevant documentation in one place.
What the validity research actually says.
The Schmidt-Hunter meta-analytic work on interview validity — along with subsequent replication studies in industrial-organizational psychology — consistently demonstrates that structured interviews with behavioral anchors have substantially higher predictive validity for job performance than unstructured interviews. The validity coefficient gap is not marginal: unstructured interviews approach the predictive value of chance for complex roles. BARS-anchored rubrics are the operational implementation of this finding.
Proofglint operationalizes this body of research. The rubric builder, blind pre-debrief scoring, and calibration view are direct implementations of what the evidence says reduces interviewer bias and improves hiring accuracy. We are not a job board, an ATS, or an AI résumé screener. We are a structured methodology layer for the part of hiring that actually predicts performance: the interview itself.
Start running structured interviews this week.
Most teams have their first rubric built within a day of getting access.