The influence of temporal distributions on causal induction from tabular data

Area of Science:

• Cognitive Psychology
• Decision Science
• Causal Inference

Background:

• Causal inference is crucial for understanding relationships in data.
• Standard models often focus on statistical contingency (e.g., P(effect|cause)).
• The role of temporal information in human causal judgment is less understood.

Purpose of the Study:

• To investigate how statistical and temporal information jointly influence causal inference.
• To determine if temporal cues affect judgments even when statistical contingency is absent.
• To challenge existing contingency-based accounts of causal induction.

Main Methods:

• Two experiments using tabular data with explicit cause-present and cause-absent conditions.
• Stimuli included relative effect frequencies (contingency information).
• Stimuli also included temporal distribution of effects (timing information).

Main Results:

• Participants integrated both statistical and temporal information in their causal judgments.
• Temporal cues (advancing/postponing effects) were assigned causal significance.
• This occurred even when the cause did not alter the overall probability of the effect.
• Observed judgments deviated from predictions of standard contingency models.

Conclusions:

• Human causal inference is sensitive to temporal dynamics beyond statistical probability.
• Temporal information plays a significant role in assigning causality.
• Existing contingency-based theories are insufficient to explain these findings.
• Future models of causal induction should incorporate temporal dynamics.