The effects of change modality on change blindness in pigeons and humans

Our experiment was a cross-species comparative study of change blindness between pigeons and humans in which we used a change detection task for humans and modified it for pigeons. This task was based on the widely used flicker task, in which an individual is presented with two similar displays in alternation and the difference between the two displays must be identified. We hypothesized that a) change detection accuracy would be lower when an inter-stimulus interval (ISI) of 80 ms was present between successive displays than when not, b) when the change was presented fewer times than many times, and c) that accuracy would be lower when a change occurred within a dimension (e.g., a color stimulus changes to a different color), than between dimensions (e.g., a color stimulus changes to a shape). Experiments were conducted on 4 pigeons and 19 humans. Repeated measures ANOVAs were conducted to analyze the data. The presence of an ISI did not appear to have a significant effect on accuracy for either pigeons or humans. We found a significant effect of repetitions for both species. We found a significant interaction between trial type (ISI, no-ISI) and number of repetitions for humans. In addition, some changes between dimensions appear to be significantly easier to detect than changes within dimensions for both species. These results suggest partial support for the presence of a change blindness effect in both species and that pigeons experience the phenomenon in a similar pattern to humans.