Computerized tasks allow a more fine-grained analysis of the strategy deployed in a task designed to map a specific ability than the usual assessment on the basis of only the level of performance. Manipulations expected to impair performance sometimes do not have that effect, probably because the level of performance alone can confound the assessment of the ability level if researchers ignore the strategy used. In a study with 1,872 participants, the authors applied the Spatial Orientation Dynamic Test-Revised (J. Santacreu, 1999) in single and dual task settings, identifying 3 different strategies. Strategy shifts were associated with the level of performance, as more apt individuals were more likely to shift to better strategies. Ignoring the strategies yields counterintuitive results that cannot be explained by simple, direct relations among the constructs involved.