Modality, probability, and mental models.

Area of Science:

• Cognitive Psychology
• Human Reasoning
• Logic and Inference

Background:

• Investigates human understanding of logical connectives, specifically 'or'.
• Examines discrepancies between formal logic and intuitive human judgments.
• Tests the predictive power of mental model theory on reasoning processes.

Purpose of the Study:

• To explore novel inference types, such as 'A or B or both' leading to 'possibly A and B'.
• To assess whether participants accept logically invalid inferences based on content plausibility.
• To contrast acceptance of invalid vs. valid inferences and compare with mental model theory predictions.

Main Methods:

• Conducted three experiments presenting participants with novel inferential tasks.
• Used plausible assertions (e.g., space tourism, antigravity materials) to frame logical problems.
• Collected judgments on inference validity and probability estimates.

Main Results:

• A majority of participants accepted logically invalid inferences ('or both' implies 'possibly and').
• Participants rejected logically valid inferences ('exclusive or' implies 'inclusive or').
• Probability estimates correlated with perceived acceptability, not strict probabilistic validity.

Conclusions:

• Findings support mental model theory's prediction of accepting plausible but invalid inferences.
• Human reasoning deviates from formal logic, prioritizing content and perceived coherence.
• Results challenge purely logic-based models of human inference, highlighting cognitive biases.