Taking a computational approach to aging: the SPAN theory of working memory

Area of Science:

• Cognitive Psychology
• Neuroscience of Aging
• Computational Modeling

Background:

• Working memory capacity declines with normal adult aging.
• The precise cause of this decline remains debated.
• A leading hypothesis links working memory decline to reduced information-processing speed.

Purpose of the Study:

• To investigate if general slowing in information processing is computationally sufficient to cause working memory decline.
• To present a production system theory of working memory (SPAN).
• To model specific tasks prominent in slowing research.

Main Methods:

• Development of a production system theory of working memory (SPAN).
• Modeling of the digit symbol task.
• Modeling of the computation span task.

Main Results:

• The SPAN models demonstrate that slowing is computationally sufficient to produce differences in working memory tasks.
• The models quantitatively match observed differences between young and older adults.
• This supports the hypothesis that general slowing underlies age-related working memory decline.

Conclusions:

• Reduced information-processing speed is a sufficient explanation for age-related working memory capacity decline.
• The SPAN theory provides a computational framework for understanding these age-related changes.
• Computational modeling is a viable tool for aging research.