Instructions The instructions asked participants to assume the role of a high-level manager working for a company that sold electronic products through company-owned retail stores. A computer-based tutorial demonstrated how to properly make two different types of accounting-based performance evaluations by navigating the information system Partici pants were informed that they must successfully complete practice problems in order to begin the compensated performance evaluation tasks. Participants were given two attempts at the practice problems and, if necessary, a tutorial was provided after the first unsuccessful attempt In both performance evaluation tasks, participants viewed four months of sales revenue and advertising expenses for their company. Participants had the option of"drilling-down into detailed sales and advertising amounts for each of the nine product/store combinations in their company(3 stores x 3 products). In one task(spatial), participants were instructed to select the product/store combination whose sales were most sensitive to advertising expenditures. In the other task(symbolic), participants were instructed to identify the oduct/store combination with the highest monthly sales total during a particular month After completing the compensated tasks, students answered demographic questions and completed the VZ-2 visualization test(Ekstrom et al. 1976; Vicente and Williges 1988 Westerman et al. 2001; Woodland and Szul 1999). The VZ-2 test was administered to control for inherent cognitive processing differences among participants. Specifically, we expected the visualization measure to serve as a covariate, controlling for potential differ ences across participants. IS-users with relatively low visualization ability, versus those with relatively high visualization ability, should have more difficulty transforming and ma nipulating relationships among variables. Thus, IS-users with relatively low visualization ability might perform more poorly when data are presented in tabular format, since they are less able to visualize the relationships among data(Woodland and Szul 1999