Related Experiment Videos
Guided search: an alternative to the feature integration model for visual search
J M Wolfe1, K R Cave, S L Franzel
1Department of Brain and Cognitive Sciences, Massachusetts Institute of Technology, Cambridge 02139.
Summary
This study found that searching for items with combined features, like color and shape, is easier than predicted by some models. Triple conjunction searches are particularly efficient, suggesting parallel processing guides attention.
Area of Science:
- Cognitive Psychology
- Visual Perception
Background:
- The Feature Integration Theory proposes serial search for conjunctions.
- Previous models struggle to explain efficient conjunction search in some cases.
Purpose of the Study:
- To investigate visual search efficiency for conjunctions of features.
- To test the predictions of the Feature Integration Theory against empirical data.
- To explore alternative models for visual search.
Main Methods:
- Participants searched sets of items for targets defined by conjunctions of color, form, orientation, or size.
- Reaction times (RT) were measured across varying set sizes.
- Data were analyzed to determine RT X Set Size function slopes.
Main Results:
- Many unpracticed subjects showed shallower RT X Set Size slopes than predicted by serial search models.
- Searches for triple conjunctions (e.g., Color X Size X Form) were more efficient than standard conjunction searches.
- Triple conjunction search efficiency was sometimes independent of set size.
Conclusions:
- The Feature Integration Theory may not fully account for all visual conjunction search data.
- A guided search model, involving parallel processing of features to guide attention, can explain the observed results.
- Multiple parallel processes enhance the efficiency of searching for triple conjunctions.