The purpose of this thesis was to (a) identify the Reinforcement Sensitivity Theory (RST) constructs that can be applied to addiction research, and (b) systematically test RST constructs with drinking determinants derived from Cox and Klinger's motivational model of alcohol use.
In the first study, age, age of drinking onset, and sensitivity to reward (SR) predicted students' alcohol use (n = 273): SR was found to be the best predictor among those drinkers who had alcohol-related personal life concerns, after controlling for the demographic variables (n = 131 ), whereas sensitivity to punishment (SP) was not related to alcohol use.
The second study partially replicated these findings, and extended them by identifying two profiles that might increase the risk of excessive drinking for affective regulation in student drinkers (n = 207). Each profile contained determinants that
were alternative representations of Cox and Klinger's motivational pathway, which terminates in the final decision to drink. Unlike Study 1, low-SP (sometimes termed "fearlessness") was related to high alcohol consumption; the relationships between SP, avoidance motivational structure, and drinking determinants were interpreted as indicators of "emotional-vulnerability" drinking (for coping and negative affect regulation reasons). The relationships between SR, approach motivational structure, and drinking determinants ( coping reasons, emotional dysregulation, enhancement and coping motives) were interpreted as indicators of "emotional-reward" drinking (for coping and positive and/or negative affect regulation reasons).
The third study tested the relationships between RST constructs and alcohol reinforcement drinking among students (n = 138) with an experimental paradigm that combined a mood-induction procedure with an alcohol-taste test. Students' SP scores
were positively related to their negative reinforcement drinking scores, and SR scores were positively related to their positive and negative reinforcement drinking scores. Study 3, failed to achieve most of its primary objects because of participant characteristics and methodological confounds. The study failed to find any support for its main hypotheses, but it still managed to identify a significant main effect for positive mood-induction condition. Participants in this condition were found to consume more alcohol during the taste-test than those participants in the negative and
neutral conditions. The findings regarding SR scores and alcohol consumption were consistent with those for Studies 1 and 2. Likewise, the findings for SP and alcohol consumption were also consistent with those reported in Study 2.
Overall, RST constructs were useful in categorising students' drinking within the framework of Cox and Klinger's model: SR scores were found to be positively related to self-reported alcohol use in each study, whereas SP scores were found to be
negatively related to alcohol use in Studies 2 and 3 only. Therefore, SR motivated behaviour is considered to be more of a risk factor for excessive drinking among young social drinkers than SP motivated behaviour. SR might be perceived to be an
aetiological factor of excessive drinking and SP a maintenance factor. The implications of these results for future research include a better understanding of the multiple determinants of drinking behaviours, risk identification, and the design of
targeted interventions.