The primary focus of this thesis was to examine the relationship of reward sensitivity and outcome expectancies, variables traditionally associated with alcohol use, to cannabis use behaviour and to explore the relationship of affect and locus of control to alcohol and cannabis use. It was hypothesised that hazardous alcohol and cannabis use would be related to higher levels of reward sensitivity and to the endorsement of more positive outcome expectancies. It was also hypothesised that positive outcome expectancies would mediate the relationship between reward sensitivity and cannabis use, and that the relationship between reward sensitivity and both alcohol and cannabis use would be moderated by punishment sensitivity. No specific hypotheses were formulated for the relationship of negative outcome expectancies, affect and locus of control of reinforcement to substance use. A total of 465 young adults aged between 18 and 35 years completed a questionnaire which assessed substance use patterns, reward and punishment sensitivity, outcome expectancies, locus of control, and affect. Results indicated that higher levels of reward sensitivity reliably distinguished hazardous from non-hazardous alcohol and cannabis users as well as cannabis users from cannabis non-users. The relationship between reward sensitivity and substance use was partially mediated by outcome expectancies, but not moderated by punishment sensitivity. An exploratory factor analysis demonstrated a high rate of concordance between alcohol and cannabis outcome expectancies. Locus of control of reinforcement was found to be unrelated to alcohol and cannabis use behaviour. There was little commonality in the relationship of sensitivity to punishment, negative outcome expectancies, and affect to alcohol and cannabis use. The second focus of this thesis was to pilot an intervention aimed at reducing alcohol use via the challenging of expectations regarding the rewarding outcomes associated with alcohol use in a group of young adult Australian males. A three session intervention was completed by three males aged between 19 and 31 years. The results demonstrated no reduction in hazardous alcohol use or global positive alcohol outcome expectancies at the completion of the intervention program or at a 3-month follow-up. Furthermore, there was no reduction in expectancies of increased sexual interest for any of the participants at the 3-month follow-up compared to baseline, despite a reduction in these expectancies for one of the three participants at the completion of the intervention. A reduction in monthly drinking levels and in expectancies of increased confidence compared to baseline was noted for two of the three participants at the 3-month follow-up. It was concluded overall that there is consistency between the relationships of reward sensitivity and positive outcome expectancies to alcohol and cannabis use and that outcome expectancies may be a proximal mechanism through which reward sensitivity influences alcohol and cannabis use. It was further concluded that whilst causal inferences regarding the effectiveness of the intervention could not be made, the results provide some evidence for the usefulness of this treatment in changing a proportion of the studied outcomes. This potentially provides an incentive for future controlled design research in larger samples and with alternate substances.

There is a pervasive belief in the United States that the college experience typically includes frequent social activities characterized by widespread alcohol use. Unfortunately, awareness of the hundreds of deaths and wide variety of other harms experienced by college students as a result of alcohol use is much less pervasive. In an effort to increase awareness of the negative impact of alcohol use on college campuses, the NIAAA commissioned a panel composed of scientists and college presidents to document alcohol-related harms and identify strategies that have been found to be effective in reducing risky alcohol use based on empirical evidence. The final report of this expert panel was released as a "Call to Action" for institutions nationwide in an effort to increase understanding of the severity and prevalence of risky alcohol use, and to provide descriptions of programs that were considered effective based on empirical evidence. Unfortunately, there were very few strategies found to be effective, and one of the effective approaches could only be implemented in specialized laboratories operated by scientists with expertise in expectancy challenge. Due to the severity and pervasiveness of the college alcohol problem and the limited number of strategies deemed effective, there is clearly a pressing need to develop and validate an expectancy challenge method that could be implemented by any institution without being limited by the need for a specialized laboratory and highly trained personnel. Achievement of these goals was the focus of the present project. To this end, an expectancy challenge curriculum designed for delivery in a college classroom was developed based on a laboratory delivered protocol previously found to be effective in reducing alcohol use among college students, and a classroom delivered curriculum previously found to be effective with high school students. The newly developed college classroom curriculum was implemented in a single session with groups of students during their regular class time in their usual classroom. Measures of alcohol use and associated harms were completed anonymously by each participant before completing the curriculum and for the month following completion of the curriculum. Analyses indicated significant reductions in alcohol consumption among males and females in comparison to students who were randomly assigned to a wait-list control condition. Unfortunately, significant reductions in alcohol-related harms were not found. The time periods for baseline and follow-up were only one month each, and that limitation in the number of opportunities to experience harms limits the likelihood of demonstrating a significant reduction in harms as well. Overall, this project represents an important advance in the development of alcohol use reduction strategies that are theory-based and effective in reducing alcohol use based on empirical evidence. In addition, the curriculum that was developed and validated in this project represents the first expectancy challenge method that can be readily implemented at any institution willing to devote one hour to reducing their students' risk for a long list of negative consequences associated with alcohol use on college campuses. Motivation and a typical classroom are all that is needed.

High-risk alcohol consumption remains a primary public health concern for students on college campuses. In response to this concern the National Advisory Council of the National Institute on Alcohol Abuse and Alcoholism created a task force to identify and recommend strategies to aid college administrators in implementing effective alcohol programming at their institutions. While most administrators report being aware of these recommendations, many have not successfully implemented empirically supported interventions on their campuses. One significant barrier is the cost and difficulty of training and hiring skilled staff to implement these interventions. Of the strategies identified as effective, challenging alcohol expectancies is the only strategy validated for group administration with college students and has significant potential to address this remaining barrier. However, current expectancy-based interventions still require highly trained expert facilitators for implementation. The present study aimed to convert the previously validated Expectancy Challenge Alcohol Literacy Curriculum (ECALC) into a digital format amenable to non-expert facilitation. The resulting digital ECALC was implemented in 48 class sections of a first year student course in a group randomized trial. It was hypothesized that receiving the digital ECALC would result in significant changes in alcohol expectancies and subsequent changes in alcohol use and related harms. Analyses revealed significant changes in both positive and negative expectancies following the digital ECALC, however no significant changes in alcohol consumption or alcohol-related harms were observed at a 30 day follow-up. Exploratory subgroup analyses revealed significant differences between experimental and control groups on average and peak drinks per sitting for classes receiving the digital ECALC during the fall semester only. Semester specific variables, environmental context, and social influence variables may have contributed to the lack of behavioral changes in the overall sample following observed expectancy changes. This study represents an important development in expectancy-based interventions for college students as the digital format removed the need for an expert facilitator and maintained significant changes in expectancies. Future studies should focus on replication of these expectancy changes and on demonstrating subsequent changes in alcohol use and related harms. The present study also represents the first evaluation of a group-administered expectancy intervention to report on intra-class correlations which will aid future researchers in designing sufficiently powered studies going forward.