In two previous experiments, Abramson and colleagues (1996) used a mirage device to demonstrate that honey bees could be influenced by a virtual reality illusion. The current study sought to replicate several key findings of their second experiment by simultaneously presenting honey bees with both a real and illusory target from which to feed. Results for sixteen subjects largely replicated those reported by Abramson and colleagues, but only when individual response patterns were considered. As a group, illusion vs. real target error rates decreased linearly over study trials. However, using organism-centered (i.e., person-centered) analyses, this effect was discovered to be attributable to only two honey bees. Considered as individuals, and consistent with previous findings, the error rates for most of the bees in this sample did not decrease over time. Results therefore support the use of the mirage device in future studies of visual illusions in insects. Implications for analyzing error rates and other important outcomes in such studies are further discussed.
Jones et al. (Sun,) studied this question.