Abstract
Little is known regarding how demographic features affect visual categorization of body size. Here, we assessed how individuals' categorization performance is impacted by gender and race. Specifically, we investigated 1) The impact of observed subjects' race and gender on observers' categorization of body images/stimuli and 2) The impact observers' race and gender had on their categorization performance. We designed an experiment to study the potential effects of these features by presenting a condition rich paradigm (24 conditions) comprised of 2 gender body stimuli (male, female), 3 race body stimuli (African-American, Caucasian, Green Avatar as a control race), and 4 categories of body mass index (BMI; underweight, normal, overweight, obese). Preliminary results demonstrated that volunteers' performance as measured by accuracy was modulated by the race and gender of the stimuli they observed. Specifically, we found a significant main effect of race of the stimuli, as categorization performance was best for the Green Avatar, followed by Caucasian and then African-American. A significant main effect of BMI weight categories was observed, as volunteers performed best for normal weight stimuli, then underweight and overweight, and worst for obese. Intriguingly, a significant interaction between gender and weight category was found. While volunteers were more accurate in categorizing male stimuli that were underweight and normal relative to overweight and obese, they were more accurate for categorizing female stimuli that were overweight and obese. We also observed a significant interaction between race, weight category, and volunteers' gender. Female volunteers were more accurate for underweight than overweight when viewing African-American stimuli. In contrast, male volunteers performed equally for underweight and overweight body stimuli. Our results illustrate that body size perception is not uniform across observers and can be influenced by demographics of both the observers and the stimuli they observe.
Meeting abstract presented at VSS 2017