Year: 2013
Type: article
Abstract: Recent work on causal learning has investigated the possible role of generic priors in guiding human judgments of causal strength.One proposal has been that people have a preference for causes that are sparse and strong-i.e., few in number and individually strong (Lu et al., 2008).Evidence for the use of sparse-and-strong priors has been obtained using a maximally simple causal set-up (a single candidate cause plus unobserved background causes).Here we examine the possible impact of generic priors in more complex, multicausal set-ups.Sparse-and-strong priors predict that competition can be observed between candidate causes even if they occur independently (i.e., the estimated strength of cause A will be lower if the strength of uncorrelated cause B is high rather than low).Experiment 1 revealed such a cue competition effect in judgments of causal strength.Experiment 2 showed that, as predicted by a Bayesian learning model with sparse-and-strong priors, the impact of the prior diminishes as sample size increases.These findings support the importance of a preference for parsimony as a constraint on causal learning.
Institution: University of California, Los Angeles
Language: English
Subfield: Artificial Intelligence
Field: Computer Science
Domain: Physical Sciences
Open Access status: green
Funder: Office of Naval Research
Award: G5196708573
OpenAlex ID: W2187275594