Effects of Level of Model Control for Implicit Learning of Decision Support Systems

Abstract

In highly complex, dynamic environments human decision making abilities can be overwhelmed by exceptionally large amounts of data and inherent uncertainty. Under these circumstances, decision support tools can significantly enhance human decision making. This experiment investigated the effect of level of model control on the development of expertise with a decision support tool through implicit learning. Two groups of participants used the decision support tool with different levels of model control (i.e., number of variables the participants are able to manipulate) across twelve training and three test sessions. Results indicated that participants learned to interact with the model for both levels of control. Participants were able to implicitly learn the interactions with no feedback and minimal instruction on the interdependencies present within the decision support model.