How shared are age-related influences on cognitive and noncognitive variables?

Area of Science:

• Cognitive psychology
• Neuroscience
• Gerontology

Background:

• Theories propose age-related cognitive decline stems from a single mechanism, like processing speed.
• Structural equation models (SEM) have supported these common factor explanations.

Purpose of the Study:

• To critically evaluate the evidence for a unitary mechanism underlying age-related cognitive decline.
• To investigate the limitations of common factor models in explaining cognitive aging.

Main Methods:

• Re-analysis of four cross-sectional datasets using structural equation modeling.
• Testing common factor and factor-specific models.
• Simulation studies with known data structures to assess model falsifiability.

Main Results:

• Common factor models fit best in three of four datasets, but significant direct age effects were also found.
• In one dataset, a factor-specific model fit as well as the common factor model.
• Simulations showed common factor models were difficult to falsify, even when incorrect, while factor-specific models were more easily falsified.

Conclusions:

• Current evidence is insufficient to conclude that all age-related cognitive declines result from a single mechanism.
• Common factor models may be overly robust and difficult to falsify with existing analytical methods, potentially masking specific age effects.