Computational Models of Cognition
Sequential Sampling Models
The Drift-Diffusion Model (DDM)
The DDM is a prominent model for 2-Alternative Forced Choice (2AFC) tasks. It conceptualizes decision-making as the noisy accumulation of evidence over time until a threshold is reached. The continuous evidence accumulation is described by a stochastic differential equation: where is the drift rate and represents Gaussian white noise.
Primary DDM Parameters
| Parameter | Psychological Interpretation |
|---|---|
| Drift Rate () | Quality or strength of information processing; task difficulty. |
| Boundary Separation () | Response caution or speed-accuracy tradeoff. |
| Starting Point () | A priori bias toward one response alternative. |
| Non-decision Time () | Time required for stimulus encoding and motor execution. |
If a participant is instructed to emphasize accuracy over speed, which DDM parameter is hypothesized to increase?
An emphasis on accuracy causes participants to widen their decision boundaries (increase boundary separation, ), requiring more evidence before committing to a choice.
Correct answer: Boundary separation
In Bayesian cognitive modeling, what term refers to the initial belief distribution before any new evidence is observed?
The prior (or prior distribution) represents the agent's initial beliefs before updating them with new data via Bayes' rule.
Correct answer: Prior
Bayesian Cognitive Modeling
Bayesian models assume the brain operates as an optimal statistical inference engine. Human cognition is modeled using Bayes' Theorem, updating prior beliefs with new evidence or likelihood to form a posterior belief :
In the Bayesian formula , what does represent?
represents the likelihood function, which is the probability of observing the data given a specific hypothesis or parameter value.
Correct answer: The likelihood function