A novel task for measuring numerical bias among adults

Bibliographic Data

ID21305809
AuthorsShachar Hochman (0000-0002-8322-3255, Ben-Gurion University of the Negev, corresponding author), Mattan S Ben-Shachar (Ben-Gurion University of the Negev), Roi Cohen Kadosh (0000-0002-5564-5469, University of Surrey), Avishai Henik (0000-0002-5665-6004, Ben-Gurion University of the Negev)
Year2026
Volume271
Pages106439
Publication date2026-06-01
Open AccessYes
TypeARTICLE
VenueCognition (JOURNAL)
Journal identifiersISSN: 0010-0277 • E-ISSN: 1873-7838
PublisherElsevier BV (PUBLISHER)
DOI10.1016/j.cognition.2026.106439
PMID41519047
OpenAlexW7119846276
LanguageEN
References cited54

Numerical bias is the spontaneous tendency to base decisions on numerical rather than equally available non-numerical information. We introduce the Congruent Learning-Incongruent Probe (CLIP) task, a computerised paradigm for indexing numerical bias in adults. The task presents digit pairs that vary in numerical value and physical size, organised into blocks. In feedback-based learning trials, digits are congruent (larger number in larger font) and participants learn which stimulus is "correct" for that block. In subsequent no-feedback probe trials (test trials), the same pairs are presented incongruently, revealing whether choices are spontaneously driven by numerical or physical dimensions. A sample of 129 adults completed a multi-day battery to validate the CLIP task. Drift-diffusion modelling indicated substantial individual differences in numerical bias. Higher numerical bias correlated positively with maths fluency and quantitative reasoning, paralleling child findings on spontaneous focus on numerosity (SFON) and maths competence. To establish convergent validity, we also administered a numerical Stroop task that requires suppressing numerical information; individuals with stronger numerical bias showed larger interference and facilitation effects. These findings validate the CLIP task as a reliable measure of numerical bias and, more broadly, highlight how variability in spontaneous numerical processing shapes cognitive-control demands, illuminating the interplay between domain-specific biases and executive function

Cognition · Computer simulation · Numerical cognition · Numerosity adaptation effect · Response bias · Stroop effect · Task (project management) · Cognitive Abilities and Testing · Cognitive and developmental aspects of mathematical skills · Numerical Analysis · Visual and Cognitive Learning Processes

Tools

Open DOIOpen Access