The validation of measures of implicit attitudes in the context of predicting disordered eating

Khan, Saira (2013) The validation of measures of implicit attitudes in the context of predicting disordered eating. (PhD thesis), Kingston University, .

Abstract

Disordered eating has a number of severe health consequences which emphasise the need for early detection to prevent further health deterioration. To gain a better understanding of the underlying thoughts which may be guiding such behaviour, social cognition research has turned to implicit measures. Drawing on this, the aims of this project were to 1) validate measures of implicit attitudes 2) explore the role of implicit and explicit attitudes in predicting disordered eating in diagnosed and healthy samples. The project addressed these aims by carrying out 4 studies. The first study validated implicit measures, the Brief Implicit Association Test (BlAT) and the Approach Avoidance Task (AAT) by investigating implicit cognitions towards food finding that the implicit measures were influenced by contextual cues and different devices. This study provided insight into the factors needed to be considered when using implicit measures in the subsequent studies of this project. In the second study, the predictive validity of the AAT used to measure automatic tendencies towards high and low fat foods alongside corresponding explicit measures for determining behavioural food choice were investigated, finding that these measures were not able to predict real-life food choices. The third study examined implicit cognitions using towards body image measured by BlATs and food measured by the AAT in a diagnosed sample of participants with an eating disorder (ED). Results showed group differences emerged on the body image implicit measures with participants with an ED showing stronger positive associations for thin body as compared to normal body images however the AAT measuring tendencies toward high and low fat foods lacked discriminating power. In the fourth and final study which administered the BlATs assessing body image, cognitive consistency between implicit and explicit attitudes related to disordered eating was explored in a general population of high school female adolescents finding a lack of consistency between the measures. Results showed a concerning prevalence of 22% reached a clinical level for ED. In addition, a majority displayed an automatic preference for thin body images regardless of their ED status. In conclusion, the results from this study demonstrate the susceptibility of implicit measures and found that implicit measures did not prove more sensitive than self¬report measures in the domain of disordered eating. Future research is recommended to further explore the usefulness of implicit measures in general and in the field of disordered eating.

Actions (Repository Editors)

Item Control Page Item Control Page