Reasoning Patterns in Solving Conjunction Fallacy Tasks

Purpose. This study examines reasoning in conjunction fallacy tasks and identifies strategies that lead to correct or erroneous responses. Design/methodology/approach . Forty participants solved five conjunction fallacy tasks: the classical Linda problem and four newly developed tasks with the same logical structure but different levels of content coherence. Content coherence refers to the extent to which the description makes the conjunctive option appear psychologically plausible. The study combined a mixed-effects logistic regression model with a phenomenological analysis of participants’ reasoning. Findings. The strongest predictor of performance was content coherence: highly coherent descriptions reduced the odds of a correct answer by approximately 25 times compared with low-coherence tasks. Phenomenological analysis showed that correct answers were associated with recognizing the relation between a general and a more specific event, critically checking the first impression, using experience as a counterexample, and relating the situation to oneself. Erroneous responses were associated with confirmation of the first impression through experience, reliance on stereotypes, and transfer of a conjunctive choice pattern across tasks. Cognitive reflection and numeracy were not significantly associated with performance. The results suggest that the crucial factor is not the mere presence of analytical resources, but whether the first intuitive impression is treated as a hypothesis to be tested or as suffi cient grounds for a decision. Research limitations/implications. The sample size was limited, and subjective process characteristics were measured after the response. Future studies should examine conditions that help participants move from intuitive impressions to testing alternative interpretations. Originality/value. The study combines statistical modeling with phenomenological analysis and shows how the conjunction fallacy emerges from the interaction between intuitive impressions, alternative mental models, and analytical verification.