Iowa Gambling Task: a preliminary comparison between different analysis strategies on a sample of Italian adolescents

The Iowa Gambling Task (IGT) is one of the most common paradigms used to assess decision-making and executive functioning. In this task, participants draw cards from four decks, each with different win/loss ratios and magnitudes. The task assesses participants’ loss aversion, memory for past payoffs, and response consistency. Traditionally, performance on the IGT is measured by the net difference between draws from 'advantageous' and 'disadvantageous' decks. However, this simple metric doesn’t consider reaction times, exploration-exploitation strategies, or speed of learning. In this study, we present two alternative methods of analysis of the IGT – the expectancy-valence model and the drift-diffusion model. In a preliminary analysis, we compared them with the difference metric using a convenience sample of 227 14-to-19-year-old Italian students (148 females; age = 16.9 ± 1.54 years) who completed a shortened version of the IGT task. We show that the alternative models allow including reaction time analysis, as well as a hierarchical parameter estimation. Results are discussed by weighing relative model performance, interpretability, ease of implementation, and estimate uncertainty, to guide researchers in choosing an approach for their studies.