From Causal Models to Sound Heuristic Inference

Ana Sofia MoraisMax Planck Institute for Human Development
Lael SchoolerMax Planck Institute for Human Development
Henrik OlssonUniversity of Warwick
Björn MederMax Planck Institute for Human Development

Abstract

Do people rely on their causal intuitions to determine the predictive value of cues? Our real-world dataset comprises one criterion (child mortality) and nine cues (e.g., GDP per capita). We elicited people’s causal models about the domain. In a second task, they had to rank the cues according to their beliefs about the cues’ predictive value. Alternative cue rankings were derived from people's causal models using measures of causal centrality. People’s judgments of cue importance corresponded more closely to the causal-based cue orders than to the statistical associations between the cues and the criterion. Computer simulations suggest that people’s causal-based cue orders form a sound basis for heuristic inference. Causal-based cue orders allowed take-the-best, a simple decision heuristic, to perform as well as a linear model using about 35% of the available data. These findings indicate that people can rely on their causal intuitions to determine a useful cue order.

Files

From Causal Models to Sound Heuristic Inference (280 KB)



Back to Table of Contents