Testing the convergent validity of the nondecision time parameter of the diffusion model

Area of Science:

• Cognitive psychology
• Computational neuroscience
• Decision-making research

Background:

• Diffusion models are used to model reaction time data in decision-making.
• Each parameter is assumed to represent a distinct cognitive process.
• Testing the validity of these parameters is crucial for accurate modeling.

Purpose of the Study:

• To test the convergent validity of the nondecision time parameter (t₀) in diffusion models.
• To investigate whether manipulations selectively affect specific diffusion model parameters.
• To examine the discriminant validity of drift rate (v) and starting point (z).

Main Methods:

• Two experiments with a total of 104 participants performing a lexical decision task.
• Manipulation of encoding difficulty to target nondecision time.
• A simulation study to analyze parameter interdependencies.

Main Results:

• Encoding difficulty significantly affected nondecision time (t₀) as hypothesized.
• Medium to large effects were also observed on drift rate (v) and starting point (z).
• Simulations suggested incorrect assumptions about intertrial variability might explain effects on drift rate.

Conclusions:

• Results support the convergent validity of the nondecision time parameter (t₀).
• Findings question the discriminant validity of drift rate (v) and starting point (z).
• Further research is needed to refine diffusion model parameter interpretation, especially concerning encoding manipulations.