Erratum: “Comparing the Iowa and Soochow gambling tasks in opiate users”

The Iowa Gambling Task (IGT) is in many respects the gold standard for demonstrating decision making in drug using groups. However, it is not clear how basic task properties such as the frequency and magnitude of rewards and losses affect choice behavior in drug users and even in healthy players. In this study, we used a variant of the IGT, the Soochow Gambling Task (SGT), to observe choice behavior in opiate users and healthy decision makers in a task where reward frequency is not confounded with the long-term outcome of each alternative. In both opiate users (n = 26) and healthy controls (n = 27), we show that reward frequency strongly influences choice behavior in the IGT and SGT. Neither group showed a consistent preference across tasks for alternatives with good long-term outcomes, but rather, subjects appeared to prefer alternatives that win most frequently. We interpret this as evidence to suggest that healthy players perform better than opiate users on the IGT because they are able to utilize gain-loss frequencies to guide their choice behavior on the task. This challenges the previous notion that poorer performance on the IGT in drug users is due to an inability to be guided by future consequences.