Using the theory of planned behaviour and implementation intentions to reduce binge drinking in new university students

Abstract

Objective: Excessive alcohol consumption, including binge drinking, increases when students enter university. This study tests whether combining messages targeting theory of planned behaviour (TPB) constructs with if-then plans (i.e. implementation intentions) to avoid binge drinking reduces binge drinking in new university students.

Design: One month after starting university, students (N = 407) were randomly assigned to condition in a 2 (TPB messages) × 2 (implementation intentions) factorial design.

Main Outcome Measures: Cognitions about binge drinking were assessed immediately post-intervention. Frequency of binge drinking was assessed at one-month follow-up (n = 205).

Results: Participants who viewed the messages had significantly weaker intentions to engage in binge drinking and less favourable cognitions about binge drinking (affective attitude, descriptive norms, and self-efficacy) than those who did not view the messages. In addition, participants who formed an implementation intention to avoid binge drinking reported significantly fewer instances of binge drinking at follow-up.

Conclusion: The findings provide some support for the use of interventions based on the TPB to reduce intentions to engage in binge drinking and for forming implementation intentions to reduce the frequency of binge drinking in new university students. No evidence was found for the synergistic effect of combining the two interventions.

Keywords:

• Heavy episodic drinking
• college
• online
• intervention
• experiment
• randomised controlled trial

Disclosure statement

No potential conflict of interest was reported by the authors.

Notes

1 In the UK, binge drinking is defined as consuming 6/8 or more units of alcohol in a single session for women/men. A unit of alcohol is equivalent to 8 grams of pure alcohol. In the USA, binge drinking is defined as consuming 4/5 or more standard drinks in a row for women/men. A standard drink is equivalent to 14 grams of pure alcohol.

2 Three additional items, rated on 7-point response scales, were included to assess the extent to which participants endorsed the three beliefs targeted by the messages. Participants who received the messages reported weaker beliefs that binge drinking would be fun, F(1, 402) = 3.99, p = .047, stronger beliefs that binge drinking would have a negative impact on their studies, F(1, 402) = 10.46, p = .001, and weaker beliefs that they would engage in binge drinking if their friends were, F(1, 402) = 4.22, p = .04, than participants who did not receive the messages. Participants also completed the measures of cognitions about binge drinking at one-month follow-up. The multivariate main effect of the messages condition on cognitions about binge drinking was significant at one-month follow-up, F(7,185) = 2.33, p = .03. Univariate F tests revealed that the main effect of the messages condition on measures of affective attitudes, F(1,191) = 3.94, p = .049, descriptive norms, F(1,191) = 8.20, p = .005, and self-efficacy, F(1,191) = 5.67, p = .02, were significant, whereas the effect on intention was non-significant, F(1,191) = 1.02, p = .31. In addition, the effect of the messages condition on the belief that binge drinking would be fun was also significant, F(1,191) = 3.24, p = .02, whereas the effect on beliefs that binge drinking would have a negative impact on their studies, F(1,191) = 1.46, p = .23, and the likelihood that they would engage in binge drinking if their friends were, F(1,191) = 1.20, p = .28, were non-significant.

3 Given the relatively high level of attrition between baseline and one-month follow-up, an intention-to-treat analysis was also conducted using last observation carried forward from baseline. The findings were unchanged. The main effects of the message condition, F(1, 400) = 0.004, p = .95, d = 0.01, and the implementation intentions condition, F(1, 400) = 2.96, p = .09, d = 0.06, were non-significant, as was the interaction between the two conditions, F(1, 400) = 0.03, p = .86. Participants were also followed-up after six months to complete the measures of alcohol consumption and cognitions again. Only 113 participants completed the six-month follow-up questionnaire (after the exclusion of three participants due to extreme levels of alcohol consumption), leading to small sample sizes in some cells (e.g. only 19 participants who received the messages and instructions to form implementation intentions completed the six-month follow-up measures). Analyses at this time point revealed that the multivariate main effect of the messages condition on cognitions about binge drinking was non-significant, as were the effects of the messages condition on the targeted beliefs. The main effects of the message and implementation intentions condition, as well as their interaction, on the frequency of binge drinking were also non-significant.