Is there a K in capacity? Assessing the structure of visual short-term memory

Area of Science:

• Cognitive Psychology
• Neuroscience
• Human Memory

Background:

• Visual short-term memory (VSTM) capacity limits are debated, with discrete-slot and continuous-resource models offering competing explanations.
• Understanding VSTM structure is crucial for cognitive goal achievement and information processing.

Purpose of the Study:

• To investigate the structure of visual short-term memory (VSTM) capacity by comparing discrete-slot and continuous-resource models.
• To assess how different VSTM models explain data from traditional VSTM tasks and generalize across tasks.

Main Methods:

• Fitting theoretical ROCs from various VSTM models to data from a change detection task (simple features) and a rapid serial visual presentation task (real-world objects).
• Conducting 3 experiments to evaluate model fit and predictive accuracy.
• Performing joint modeling analyses to assess model performance across both tasks.

Main Results:

• Consistent support was found for pure continuous-resource models of VSTM across experiments.
• Joint modeling analyses reinforced support for continuous-resource models but indicated task-specific parameters were necessary.
• Signal detection model interpretations suggest memory signal variations due to memoranda and encoding conditions.

Conclusions:

• Pure continuous-resource models provide the best account of visual short-term memory capacity.
• While a common parameter set across tasks is not supported, the findings align with a graded resource allocation view of VSTM.
• Variability in memory signals, influenced by item differences and encoding, is key to understanding VSTM limits.