The Cognitive Processing of Causal Knowledge

TL;DR

This paper explores how human causal reasoning aligns with Bayesian network models, supported by psychological studies on discounting and Piaget's observations, suggesting humans learn and reason about causality similarly to algorithms.

Contribution

It links human causal reasoning and learning to Bayesian network algorithms, proposing a subjective causality definition and discussing methods for empirical validation.

Findings

Psychological studies support Bayesian inference in human reasoning.

Humans learn causal structures by observing dependencies similar to algorithms.

Proposes methods to empirically test the alignment between human and algorithmic causal reasoning.

Abstract

There is a brief description of the probabilistic causal graph model for representing, reasoning with, and learning causal structure using Bayesian networks. It is then argued that this model is closely related to how humans reason with and learn causal structure. It is shown that studies in psychology on discounting (reasoning concerning how the presence of one cause of an effect makes another cause less probable) support the hypothesis that humans reach the same judgments as algorithms for doing inference in Bayesian networks. Next, it is shown how studies by Piaget indicate that humans learn causal structure by observing the same independencies and dependencies as those used by certain algorithms for learning the structure of a Bayesian network. Based on this indication, a subjective definition of causality is forwarded. Finally, methods for further testing the accuracy of these claims are discussed.