No abstract

No abstract

Three experiments examined the effects of interactive visualizations and spatial abilities on a task requiring participants to infer and draw cross sections of a three-dimensional (3D) object. The experiments manipulated whether participants could interactively control a virtual 3D visualization of the object while performing the task, and compared participants who were allowed interactive control of the visualization to those who were not allowed control. In Experiment 1, interactivity produced better performance than passive viewing, but the advantage of interactivity disappeared in Experiment 2 when visual input for the two conditions in a yoked design was equalized. In Experiments 2 and 3, differences in how interactive participants manipulated the visualization were large and related to performance. In Experiment 3, non-interactive participants who watched optimal movements of the display performed as well as interactive participants who manipulated the visualization effectively and better than interactive participants who manipulated the visualization ineffectively. Spatial ability made an independent contribution to performance on the spatial reasoning task, but did not predict patterns of interactive behavior. These experiments indicate that providing participants with active control of a computer visualization does not necessarily enhance task performance, whereas seeing the most task-relevant information does, and this is true regardless of whether the task-relevant information is obtained actively or passively.

We examined changes in performance as people learned to use an angled laparoscope, a challenging spatial skill that must be mastered by surgeons who perform minimally invasive techniques. In Experiment 1, novices took tests of spatial and general reasoning ability, and then learned to operate an angled laparoscope, simulated in a virtual environment, over 12 learning sessions. Initial performance showed considerable variability among learners, with performance related to general and spatial abilities. As learning progressed, interindividual variability diminished and all learners attained proficiency; the correlation with general ability diminished but the correlation with spatial ability remained significant. In Experiment 2, performance by highly experienced surgeons on the simulation was excellent from the first session, confirming its ecological validity. The findings contribute to theories of skill acquisition. They also have practical implications for the selection of surgeons and for the potential use of virtual environments in surgical training.