--- - |2 Abstract The Iowa gambling task (IGT) was developed as a neuropsychological assessment of “real‐life” decision making in a laboratory setting by Bechara, Damasio, Damasio, and Anderson in . The IGT was originally implemented as a manual task during which the participants selected paper cards. The task has since evolved into a computerized assessment tool (Psychological Assessment Resources IGT or PAR™IGT) that uses only virtual cards and reinforcement schedules that are somewhat different from those in the original version. Caution should be used when psychological assessments change from manual to virtual formats. Such changes may unintentionally introduce confounds that render the 2 tests nonequivalent as shown with the Wisconsin Card Sorting Task by Steinmetz, Brunner, Loarer, and Houssemand in . In this study, we compared IGT performance when participants used either a laptop computer or a touchscreen tablet. Each device was used in 2 conditions: “involved” and “noninvolved.” Participants in the involved condition turned over physical cards and mimicked their own selections by clicking (laptop) or reaching out and touching (tablet) the corresponding virtual cards. Participants in the noninvolved condition only made selections of virtual cards, whereas an adjacent experimenter turned over the corresponding physical cards. Results showed that performance systematically and significantly improved as degree of physical involvement increased in the following order: laptop noninvolved (lowest performance), tablet noninvolved, laptop involved, and tablet involved (highest performance). In addition, participants' verbalized understanding of the task was significantly more accurate with increased physical involvement. - Journal of Behavioral Decision Making, EarlyView.