The Capacity of Visual Short-Term Memory is Set Both by Visual Information Load and by Number of Objects

Alvarez, George A.; Cavanagh, Patrick · 2004 · OpenAlex-citations

DOI: 10.1111/j.0963-7214.2004.01502006.x

archive: archived pipeline: cataloged verified

Get this paper ↗ (DOI — opens at the source; we link to it, we don't host it)

Summary

This study investigates the determinants of visual short-term memory (VSTM) capacity, challenging the prevailing view that VSTM is limited solely by a fixed number of objects (approximately four), regardless of their complexity. Previous research by Luck and Vogel (1997) suggested that memory capacity is object-based rather than feature-based. Alvarez and Cavanagh reexamine this by testing whether VSTM capacity is instead constrained by the total amount of visual information that can be stored. They hypothesize that if there is a fixed information limit, the number of objects stored should vary inversely with the visual information load per object. To test this, the authors employed two behavioral measures across five stimulus classes: colors, letters, Chinese characters, random polygons, and shaded cubes. Visual search rate was used as an index of the visual information load per object, under the assumption that more complex objects require slower processing. VSTM capacity was estimated using a change detection task, determining the maximum number of objects participants could retain. The study included six participants who completed multiple sessions. Control experiments were conducted to rule out alternative explanations, such as encoding time limits, visual similarity effects, or low-resolution storage slots. The results demonstrated that VSTM capacity varied significantly across stimulus classes, ranging from 1.6 objects for shaded cubes to 4.4 for colors. Crucially, there was a strong linear relationship ($r^2 = .992$) between visual search rate and the reciprocal of memory capacity. As the information load per item increased (indicated by slower search rates), the number of items that could be stored decreased. Extrapolating this linear relationship revealed an upper bound on capacity of approximately four to five objects, even for items with minimal information load. Control experiments confirmed that these findings were not due to insufficient encoding time or visual similarity, and that single complex objects were stored with high precision, ruling out low-resolution slot models. The authors conclude that VSTM capacity is set by two simultaneous limits: a total visual information load and a maximum number of objects. While the total amount of visual detail stored determines how many objects can be held, there is an absolute ceiling of four or five objects, regardless of how simple those objects are. This finding reconciles previous conflicting results by suggesting that basic features may be stored in separate channels, while complex objects require more information load, reducing the total number of items that fit within the fixed information budget. The study implies that VSTM is flexible in allocating resources based on object complexity but is ultimately bounded by both informational and numerical constraints.

Provenance

The full processing record for this entry. Every stage of this paper's journey through the pipeline is logged — what ran, with which tool and model, how many attempts it took, and when it last completed.

Summary generated by qwen3.6-27b-prismaquant on 2026-06-26; verification: verified.

Topics

Ranked by relevance to this paper. Hover a topic for its definition.

• working memory