A diffusion model decomposition of the effects of alcohol on perceptual decision making

Even in elementary cognitive tasks, alcohol consumption results in both cognitive and motor impairments (e.g., Schweizer and Vogel-Sprott, Exp Clin Psychopharmacol 16: 240-250, 2008).

The purpose of this study is to quantify the latent psychological processes that underlie the alcohol-induced decrement in observed performance.

In a double-blind experiment, we administered three different amounts of alcohol to participants on different days: a placebo dose (0 g/l), a moderate dose (0.5 g/l), and a high dose (1 g/l). Following this, participants performed a "moving dots" perceptual discrimination task. We analyzed the data using the drift diffusion model. Model parameters drift rate, boundary separation, and non-decision time allow a decomposition of the alcohol effect in terms of their respective cognitive components, that is, rate of information processing, response caution, and non-decision processes (e.g., stimulus encoding, motor processes).

We found that alcohol intoxication causes higher mean RTs and lower response accuracies. The diffusion model decomposition showed that alcohol intoxication caused a decrease in drift rate and an increase in non-decision time.

In a simple perceptual discrimination task, even a moderate dose of alcohol decreased the rate of information processing and negatively affected the non-decision component. However, alcohol consumption left response caution largely intact.