An Exploration of Explicit and Implicit Attitudes, Preferences and Anticipated Behavioural Responses Towards Anorexia Nervosa and Obesity

Previous explicit mental health and weight bias studies have suggested that blame is attributed towards both anorexia nervosa (AN) and obesity sufferers for their conditions. Implicit weight bias research has consistently supported a preference towards underweight compared to overweight individuals. The present study aimed to examine the extent to which this preference existed when comparing the more extreme weights associated with AN and obesity, assessed using both explicit and implicit measures. Fifty-four undergraduate psychology students were recruited for the present experiment. After reading vignettes, participants were instructed to complete the Attribution Questionnaire (Corrigan, Markowitz, Watson, Rowan, & Kubiak, 2003) and the Social Distance Scale (Breheny, 2008) as the explicit measures of attitude and anticipated behavioural responses. Participants were then required to complete two Implicit Association Tests. Participants were instructed to group together different categories of words. The time taken to complete the categorisations is designed to reflect participant preference for one of the categories (Greenwald, McGhee, & Schwartz, 1998). The present experiment was a repeated measures design: all participants completed both implicit and explicit measures when considering both AN and obesity. The results from the explicit measures suggested a significantly higher level of blame was attributed towards the obese target when compared to the AN target. Significantly more pity was reported towards AN relative to obesity. In terms of the implicit measures, nonsignificant differences were found in response latencies between AN and obesity, indicating the lack of an implicit preference towards either condition. The implicit and explicit findings were correlated but no moderate to strong significant relationships were found, suggesting inconsistencies between the implicit and explicit measures of attitude and anticipated behavioural responses. This inconsistency between measures was consistent with previous research and is discussed in relation to the Model of Dual Attitude (Wilson, Lindsey, & Schooler, 2000).