Pseudocontingency inference and choice: The role of information sampling

Area of Science:

• Cognitive Psychology
• Decision Making
• Behavioral Economics

Background:

• Pseudocontingencies are inferred correlations based on skewed base rates across a third variable.
• Previous research primarily used predetermined learning trials to study pseudocontingency effects.
• Understanding pseudocontingency's impact on choice behavior is crucial for predicting decision-making.

Purpose of the Study:

• To investigate the effect of pseudocontingency inference on choice behavior.
• To compare the role of free information sampling versus predetermined learning in pseudocontingency effects.
• To examine pseudocontingency inference in bivariate and trivariate scenarios.

Main Methods:

• Replication of previous findings on pseudocontingency effects in choice behavior (Experiment 1).
• Comparison of predetermined information versus free information sampling in a bivariate decision scenario (Experiment 2).
• Investigation of pseudocontingency inference through context-based and context-and-option-based information sampling in trivariate scenarios (Experiments 3 & 4).

Main Results:

• Pseudocontingency effects were replicated.
• An asymmetry was observed between positive (gains) and negative (losses) contexts.
• In positive contexts with self-determined information sampling, participants preferred the predominant option, indicating suboptimal choices.

Conclusions:

• Self-determined information sampling significantly influences pseudocontingency effects on choice behavior.
• The findings highlight the impact of context (gains vs. losses) on decision-making under pseudocontingency.
• Understanding these biases is essential for designing more effective decision support systems.