Causal judgement from contingency information: judging interactions between two causal candidates

Area of Science:

• Cognitive Psychology
• Causal Inference
• Decision Making

Background:

• Understanding how individuals attribute causality is fundamental to cognitive psychology.
• Previous research has explored various factors influencing causal judgments, including contingency and covariation.

Purpose of the Study:

• To investigate how people evaluate the likelihood of an interaction between two causal candidates (A and B) leading to an effect.
• To determine which types of evidence (confirmatory vs. disconfirmatory) and occurrence rates most influence these interaction judgments.

Main Methods:

• Two experiments were conducted where participants judged the extent to which an effect could be attributed to an interaction between two causal candidates.
• Experiment 1 manipulated the proportion of confirmatory instances while holding objective contingency constant.
• Experiment 2 manipulated the occurrence rate of the effect under different combinations of candidate presence/absence.

Main Results:

• In Experiment 1, judgments were influenced by the proportion of confirmatory instances, with positive instances (both causes present, effect occurs) being more influential than negative instances (both causes absent, effect absent).
• In Experiment 2, the occurrence rate of the effect when both candidates were present most strongly determined interaction judgments.
• The occurrence rate when both candidates were absent also had a significant effect, but isolated candidate presence had no significant impact.

Conclusions:

• Causal judgments, particularly regarding interactions, appear to be guided by a general model where individuals assess the proportion of instances supporting a given interpretation.
• The findings highlight the importance of specific evidential patterns and occurrence rates in shaping attributions of interactive causality.