Effectiveness of experimenter-provided and self-generated implementation intentions to reduce alcohol consumption in a sample of the general population: A randomized exploratory trial.

OBJECTIVE: To test the effectiveness of implementation intentions to decrease alcohol consumption and control for possible demand characteristics by employing an active control condition and contrasting experimenter-provided with self-generated implementation intentions. DESIGN: Two hundred forty-eight participants were randomly allocated to 1 of 4 conditions: questionnaire-only; questionnaire plus planning instruction; questionnaire, planning instruction plus experimenter-provided implementation intention; or questionnaire, planning instruction plus self-generated implementation intention. MAIN OUTCOME MEASURE: Alcohol intake. RESULTS: There were clinically and statistically significant decreases in alcohol consumption in the 2 experimental conditions, but not in the 2 control conditions, F(3, 237) = 3.34, p < .05, etap(2) = .04. There were no significant differences between experimenter-provided and self-generated implementation intentions (p = .62). Compliance moderated the effects of self-generated implementation intentions only, such that alcohol intake only significantly decreased in participants who complied with the instructions, F(1, 52) = 4.20, p < .05, etap(2) = .07. However, simply choosing an experimenter-provided implementation intention was just as effective as writing it out in full, implying that implementation intentions work even with minimal information processing. CONCLUSION: The findings further support use of implementation intentions to protect against health risk behaviors and are congruent with laboratory research showing that implementation intentions are a case of strategic automaticity.