Computer Adaptive Testing (CAT) across 9 cognitive game types: pattern recognition, working memory, task switching, processing speed, spatial reasoning, prioritization, ethical dilemmas, attention to detail, numerical reasoning. 95+ calibrated items. Difficulty adjusts per answer.
Classic psychometric tests take 60+ minutes and bore strong candidates into dropping out. HRBlade's 9 adaptive games run in 10 minutes — difficulty rises when the candidate gets it right, drops when they don't, so you get a precise percentile in a fraction of the time.
Pattern recognition, working memory, task switching, processing speed, spatial reasoning, prioritization, ethical dilemmas, attention to detail, numerical reasoning.
Each correct answer raises the difficulty band; each wrong one lowers it. The model converges on the candidate's true ability quickly — no need for 60-question tests.
Item bank validated against psychometric norms with item response theory (IRT). Difficulty parameters re-calibrated quarterly as the candidate pool grows.
Candidates are ranked against a global cohort of 100k+ test-takers, segmented by role family. Your hire's score is meaningful in context, not just a raw number.
Tab-switch detection, time-per-item anomaly flags, IP/device fingerprinting, optional webcam proctoring. Cheaters land in a flagged-for-review queue.
Whole battery completes in ~10 minutes vs. 45–60 for classical tests. Drop-off rate falls from ~30% to under 8% in our pilot data.
Engineering roles get pattern recognition + working memory + processing speed; sales roles get prioritization + ethical dilemmas + attention to detail. Configurable per requisition.
Game UI is mobile-friendly and instructive. Difficulty self-adjusts. The AI tracks not just correctness but reaction time, retry patterns and confidence signals.
Result page shows percentile vs. cohort, 5 competency scores, recommended pass/fail decision and follow-up interview probes for the weak areas.
Source: HRBlade product spec, 2025
Source: Internal benchmark across 12k candidates, 2025
Source: Customer-reported A/B benchmark, 2025
Customer support, retail, sales — when CV signals are thin, cognitive assessments are the strongest predictor of on-the-job performance. Run on 100% of applicants.
Pair pattern recognition + working memory + spatial reasoning with a take-home coding test. Cognitive scores correlate with code review pass-rates at r = 0.62.
Ethical dilemmas + prioritization + processing speed predict managerial promotions in our 6-month follow-up cohort with 83% accuracy.
Items are designed to be culture- and language-neutral (no idioms, no culture-specific references). We benchmark adverse impact across protected groups annually with a third-party I/O psychologist; results published on request.
Multi-agent AI simulates real team dynamics — before you hire, promote, or reorganize.
Learns from your hiring history. Correlation between AI score and 6-month performance: r = 0.74.
Async and live. Auto-transcription, 5-competency scoring, multi-layer anti-cheat detection.
