Similarity, not complexity, determines visual working memory performance.

Area of Science:

• Cognitive Psychology
• Neuroscience

Background:

• Visual working memory (WM) performance is often poorer for complex items, attributed to information load.
• Alternative theories propose perceptual similarity, not complexity, drives WM deficits.
• Previous studies lack control for stimulus category, confounding complexity and similarity effects.

Purpose of the Study:

• To investigate the independent effects of complexity and similarity on visual working memory.
• To clarify whether complexity or similarity is the primary determinant of WM performance.
• To examine these relationships within a controlled stimulus category (abstract polygons).

Main Methods:

• Used a delayed discrimination task to assess WM for complex versus simple abstract polygons.
• Manipulated the number of simultaneously presented items (1-4).
• Varied the perceptual similarity between the test item and the encoded sample items during retrieval.

Main Results:

• WM performance was significantly poorer for complex items compared to simple items, but only when the test item was perceptually similar to an encoded item.
• No significant difference in WM performance between complex and simple items was observed when the test item was dissimilar or identical to encoded items.
• These findings indicate that similarity at retrieval, rather than item complexity, is the critical factor.

Conclusions:

• The complexity effect in visual working memory is better explained as a similarity effect.
• The retrieval stage plays a crucial role in governing working memory performance.
• Findings necessitate a re-evaluation of current models of working memory capacity and limitations.