Two experiments compared real and virtual models as aids for learning assembly skills. In Experiment 1, ten participants individually studied either a fully assembled model, or a computer-generated one, in exploded view, that could be spatially manipulated in any direction. Participants then assembled the object in front of a video camera. ANOVA indicated virtual model are studied significantly longer but yield faster assembly than a real model. Experiment 2 used the same treatments plus a fully assembled virtual mode, randomly assigned to 28 participants who studied the aid, assembled the model, and then repeated the task from memory 3 days later. ANOVA indicated no differences between the three groups in assembly speed or accuracy. However, participants studied the exploded virtual model significantly longer than the two intact views of the model suggesting the former may impose a greater cognitive load due to the additional visual information it provides.
