ATTENTIONAL BIAS FOR ALCOHOL RELATED CUES: AN EXPLORATION OF ITS STATE-MOTIVATIONAL NATURE IN THE REAL WORLD

Incentive motivational models of addiction suggest that alcohol-related cues would automatically attract and hold attention (attentional bias - AB). Although AB is expected to be associated with alcohol-use behaviours, results in the literature have been equivocal. One explanation might be that AB is a state variable and, therefore, its magnitude and its association with alcohol-use behaviour will depend on the motivational context in which measures are taken (the state-hypothesis of AB). The aim of the thesis was to examine the state-motivational nature of AB in the real world. For that purpose, a smartphone application (app) with the alcohol Stroop-task was developed. First, the thesis examined the psychometric properties of AB measures taken on the app in the real world (chapter three), and compared them with measures taken on different devices (e.g. computers) and on different environments (e.g. neutral laboratories and a Bar laboratory). The app was then used in a series of ecological momentary assessments (EMA) to examine the effect of the motivational context on magnitude of AB and on the association between AB and alcohol consumption. In the studies described in chapter four, AB and craving were measured in close temporal proximity to alcohol consumption, as well as in non-proximal assessments. In chapter five, AB and craving were assessed in moments of temptation to drink alcohol, in moments when alcohol was present and readily available for consumption (total availability of alcohol), and in fixed control moments. Finally, a focus group was conducted to explore attitudes of prospective users of the app (chapter six). Measures of AB taken on the app in the real world showed good internal reliability but poor predictive validity. All other administration modalities showed poor psychometric properties. Results from the EMAs showed that AB did not increase in magnitude when measured proximally to alcohol consumption, in moments of temptation, or in moments of total availability. In one study, AB predicted alcohol consumption when measured proximally to alcohol use but this was not replicated in a second study. AB did not predict alcohol consumption when measured in moments of temptation or of total availability. Craving increased in magnitude when measured prior to initiation of drinking, in moments of temptation, and in moments of total availability of alcohol. Craving predicted alcohol consumption when measured proximally to alcohol use and in moments of total availability, but not in moments of temptation. However, across all EMAs, measures of craving across the day were not associated with measures of AB. Findings from the focus group revealed that attitudes of prospective users of the app could be organized in three themes; users as participants, users as clients, and users as researchers. Re-development of the app in the future should focus on promoting autonomy of use, effectiveness of the app, and motivation for engagement. Taken together, results did not support the state-hypothesis of AB and suggested that AB might have limited utility in predicting alcohol consumption. Craving increased in all situations when motivation to drink was expected to be high, and it robustly predicted alcohol consumption in most occasions. Future programs aiming to identify in real time moments of increased likelihood to consume alcohol should probably focus more on craving, rather than AB. The thesis also demonstrated the importance of measuring AB and craving in the real world, and that smartphone applications can be effectively used for that purpose. However, app developers should take attitudes of prospective users into account and balance the scientific validity of the app with its usability, user-friendliness and attractiveness