Statistical learning creates implicit subadditive predictions
Skip to main content
eScholarship
Open Access Publications from the University of California

Statistical learning creates implicit subadditive predictions

Abstract

The cognitive system readily learns when multiple cues jointly predict a specific outcome. What is less known is how the mind generates predictions when only a single cue is present. In four experiments, participants were first exposed to two objects followed by a circle with a specific size or a specific numeric value. Afterwards, participants viewed a single object and estimated the associated size or value. Finally, participants recalled the size or value that followed the initial two objects. We found that the estimated size associated with the single object was significantly smaller than 100% but significantly larger than 50% of the recalled size associated with the two objects. No participants were consciously aware of the associations. The results reveal a new consequence of statistical learning on automatic inferences: When multiple objects were previously associated with an outcome, the single object is implicitly expected to predict a subadditive outcome.

Main Content
For improved accessibility of PDF content, download the file to your device.
Current View