Expectations and interpretations during causal learning.

Area of Science:

• Cognitive Psychology
• Causal Inference

Background:

• Traditional causal induction models assume uniform interpretation of covariation information.
• Existing models do not account for dynamic, expectation-based interpretation of data during learning.

Purpose of the Study:

• To investigate how expectations influence the interpretation of covariation information in causal learning.
• To examine the impact of dynamic interpretation on causal strength judgments.

Main Methods:

• Experiments 1A-1C involved participants learning causal relationships through observed events.
• Experiment 2 manipulated the type and order of covariation information presented.
• Participants made causal strength judgments based on their learning experiences.

Main Results:

• Participants' interpretations of observations were dynamic, expectation-based, and influenced subsequent causal judgments.
• Contrary to traditional models, joint presence/absence trials could decrease causal judgments.
• Conversely, trials with one event in the absence of another could increase causal judgments.

Conclusions:

• Learner expectations dynamically shape the interpretation of covariation data.
• Findings challenge existing causal induction models and support top-down, Bayesian approaches.
• Future research should consider the role of prior expectations in causal learning.