Priming by the mean representation of a set

Area of Science:

• Cognitive psychology
• Visual perception
• Statistical learning

Background:

• Human observers can represent sets of objects using statistical properties like mean size.
• Previous research indicates that probes matching the mean size of a set are more likely to be recognized as belonging to that set.

Purpose of the Study:

• To provide further evidence for set representation by statistical properties.
• To investigate priming effects based on the mean size of a set of circles.

Main Methods:

• Participants viewed sets of circles followed by a degraded target circle outline.
• Observers judged the contrast of the target circle.
• The target circle's size was manipulated to match either the mean size or a member size of the preceding set.

Main Results:

• Target contrast was significantly higher when the target circle's size matched the mean size of the preceding set.
• Priming by the mean size of the set was stronger than priming by an individual set member.

Conclusions:

• Conceptual priming by a statistical summary description can be more potent than exact repetition priming.
• This study supports the idea that the brain extracts and uses statistical information from sets of objects for perception.