Assessing judges' use and awareness of cognitive heuristic decision-making

Abstract

Heuristics, characterized as concise cognitive shortcuts rooted in intuitive reasoning, are both capable of facilitating swift judgments and cognitive efficiency, but also introducing cognitive biases during decision-making. The judicial domain, renowned for its demanding decision-making processes, is an interesting field for studying heuristics. In this study, we developed a novel Judicial Heuristics Assessment Questionnaire (J-HAQ) and administered it to a sample of 52 judges (20 males, Mage = 45.50, SD = 8.10), with active duty in various courts across Greece. We also evaluated their analytical System 2 thinking skills using the Cognitive Reflection Test (CRT). This research pursued three objectives: (a) to explore the psychometric properties of the J-HAQ; (b) to investigate the correlation between judges' perceived use of heuristics/metacognitive awareness and their objective performance on reflective thinking; (c) to assess the correlation of self-reported usage of different heuristics and explore the influence of judges' demographics (educational level, gender, age, and years of experience) in the utilization of the reported heuristics in decision-making. Findings from a Principal Component Analysis on J-HAQ scores revealed four distinct factors (Availability, Confirmation Bias, Representativeness, and Anchoring) demonstrating sufficient reliability. We also report a significant correlation between CRT scores and reported use of the anchoring heuristic (ρ = 0.29, p = 0.04). Finally, we discovered two clusters defined by different awareness of the use of various heuristics, as well as significant association of educational level with this usage. Despite the limitations of a relatively small sample size, these findings reveal a dynamic for further interesting results from research in this domain.

Cite this work

No public copy is stored for this record, so only the metadata is cited.

Stored only when both papers are in the corpus — 2 citations exist upstream.

Papers that cite the same references — the relational fallback while embeddings are still backfilling.