RESEARCH NOTES · 5 SEPTEMBER 2026
Research-informed.
Open about the limits.
HiringScope draws on research into structured assessment and work samples. These publications inform design choices; their authors have not evaluated or endorsed this product. HiringScope has not yet been validated as a predictor of job performance.
Structure before intuition
A major review examines interview structure, rating scales, bias, follow-up questions and applicant reactions.
Our application: HiringScope lets assessors fix questions, weights and scoring anchors before invitations. A written task workflow is not equivalent to a validated structured interview.
Levashina et al. (2014) · Personnel Psychology ↗Assess work that resembles the role
A meta-analysis found work samples related to job performance, while revising earlier, higher estimates downward.
Our application: Use realistic, role-relevant tasks. Our short written scenarios are only samples of evidence; they do not directly demonstrate spreadsheet, presentation or interpersonal execution.
Roth, Bobko & McFarland (2005) · Personnel Psychology ↗Keep predictive claims proportionate
Revised meta-analytic estimates reduced the apparent predictive validity of many selection methods; structured interviews ranked highest in that comparison.
Our application: Our 0–100 task score is not an IQ, a percentile or a probability of succeeding at work. This research does not establish HiringScope’s predictive validity.
Sackett et al. (2022) · Journal of Applied Psychology ↗Context matters
The authors discuss variation between studies and the implications of revised validity estimates for selection-system design.
Our application: Treat role, language, assessment conditions and scoring version as part of the evidence. Scores from different custom assessments are not automatically comparable.
Sackett et al. (2023) · Industrial and Organizational Psychology ↗Personality research is not a personality diagnosis
A broad synthesis links conscientiousness with occupational outcomes and examines how its relationship with performance varies by context.
Our application: We assess the steps described in a response, such as checking a discrepancy. We do not infer a stable personality trait from a few chatbot answers.
Wilmot & Ones (2019) · PNAS ↗Self-report does not establish honesty
A meta-analysis of 20 within-person studies found more socially desirable personality responses in high-stakes applicant settings.
Our application: HiringScope does not certify honesty or authorship. Use clear assistance rules and an observed follow-up to explore how an answer was produced.
Hu & Connelly (2021) · International Journal of Selection and Assessment ↗Neuroscience needs a direct connection
An experiment examined how irrelevant neuroscience information can affect judgments of psychological explanations.
Our application: We do not use brain-based branding to imply validation. HiringScope measures no neural activity; a neuroscience citation alone cannot validate its scoring.
Weisberg et al. (2008) · Journal of Cognitive Neuroscience ↗What our pilot can—and cannot—show
Feedback can tell us about usability, clarity and perceived usefulness. It cannot establish that a score predicts future performance. That would require a separately designed study, appropriate participation arrangements, consistent assessments, independent scoring and relevant outcomes over time.
These are selected research notes, not a systematic review. Some findings were checked against publisher or PubMed abstracts rather than full papers. Before stronger product claims, we plan a fuller evidence review and assessment-specific validation with organisational psychology expertise.
Back to HiringScope →