Abstract
The human brain often processes visual information from different categories in the environment to create a coherent percept. For both visual objects and texts, subjective preferences for curved shapes and particular fonts have been consistently observed. Do the preferred qualities of visual properties for objects and texts have similar or different influences on subjective preference and attention allocation for visual displays that are consisted of these categories? We combined simple abstract objects and 4-letter nonsense words (e.g., galu, qita) to form novel object-text displays. The objects were either curved or sharp-edged; the words were presented in either preferred or non-preferred fonts. A pilot study showing individually presented objects or words confirmed the differential subjective preferences between curved vs. sharp-edged shapes and preferred vs. non-preferred fonts. In the main experiment, participants (n=24) provided subjective preference ratings to each object-text display (“logo”), while their eye movements were recorded for 2 seconds from the stimulus onset. Critically, we found different effects of preferred vs. non-preferred qualities of shape and font on both subjective preference and eye movements. For subjective preference, higher ratings were found with preferred than non-preferred shapes, and longer response times were found with non-preferred than preferred fonts. For eye movements, while initial attention, as indicated by the first fixations, was more likely directed to word than object components, and was also more likely drawn towards preferred than non-preferred fonts, the first fixations were shorter on word than object components. More importantly, while participants viewed the object and text components for comparable amounts of time, longer viewing time and more fixations were found for preferred than non-preferred shapes, but also for non-preferred than preferred fonts. These results suggest that the preferred qualities of visual properties for objects and texts play different roles on subjective preference and viewing patterns for multi-component displays.