Masserang, Kathleen, Nicola Mucci, Harold Greene, V. Barry Dauphin, and Remy Ford
The fundamental problem of vision is that the retinal image has countless possible interpretations in 2D and 3D space. Despite this, humans usually tend to agree on visual interpretations. In some instances, it is likely that we use heuristics that stem from regularities in our perceptual environment to constrain our interpretations. For instance, faces, which are important for identification, and for judging intentionality are usually near the top of animals. As well, the tops of objects are usually better illuminated (given the regularity of overhead light sources), and therefore, more informative than the bottoms. Finally, the slope of the visual world away from the observer places objects in extra-personal space towards the upper visual field.
Given environmental regularities, do we learn to weight attention more heavily towards object tops than bottoms? In the preliminary experiment we report here, Rorschach cards were digitized and presented with initial eye position centered on the digitized images. Rorschach images are particularly useful for the present experiment because they are initially ambiguous, and become less ambiguous as the observer imposes form. We hypothesized that if subjects have a bias to attend more to object tops than bottoms, then eye fixation patterns would be weighted towards the top. The data were subjected to a 2 (Image orientation: Standard, Inverted) X 3 (Scan phase: Early, Late, Very late) X 2 (Exposure: First, Second) X 2 (Region of interest: Top, Bottom) ANOVA, and the results supported our hypothesis. This work provides evidence of a top bias in overt allocation of attention irrespective of scan phase, exposure, and image orientation.