The diffusion model's drift rate parameter primarily reflects efficiency, rather than speed, of evidence accumulation

Area of Science:

• Cognitive psychology
• Psychometrics
• Computational neuroscience

Background:

• The diffusion decision model (DDM) drift rate (v) parameter is linked to cognitive individual differences and higher-order functioning.
• v is often interpreted as processing speed, but this mapping is debated.
• Racing accumulator models offer a more nuanced view of evidence accumulation.

Purpose of the Study:

• To investigate whether the DDM's v parameter reflects speed of evidence accumulation (SEA) or efficiency of evidence accumulation (EEA).
• To compare DDM and racing accumulator models in accounting for cognitive individual differences.
• To re-evaluate the interpretation of v in relation to processing speed and cognitive abilities.

Main Methods:

• Applied DDM and racing accumulator models to three tasks across three independent datasets.
• Analyzed the relationship between the v parameter and measures of SEA and EEA.
• Correlated v, SEA, and EEA with measures of general cognitive ability, working memory, and executive function.

Main Results:

• Across all tasks, the v parameter was more strongly related to EEA than SEA.
• EEA consistently correlated with general cognitive ability, working memory, and executive function.
• SEA explained less than 1% of the variance in cognitive measures.

Conclusions:

• Individual differences in the DDM's v parameter primarily reflect EEA, not SEA.
• The common practice of equating v with traditional processing speed is challenged.
• Findings support the link between evidence accumulation efficiency and higher-order cognitive abilities.