Interaction effects in the theory of planned behaviour: Predicting fruit and vegetable consumption in three prospective cohorts

Abstract

Objective The theory of planned behaviour (TPB) has been criticized for not including interactions between major constructs thought to underlie behaviour. This study investigated the application of the TPB to the prediction of fruit and vegetable consumption across three prospective cohorts. The primary aim of the study was to investigate whether interactions between major constructs in the theory would increase the ability of the model to predict intention to consume fruit and vegetables (i.e.; attitude × perceived behavioural control [PBC], subjective norm × PBC, subjective norm × attitude) and self-reported fruit and vegetable intake (i.e.; PBC × intention). Design Secondary data analysis from three cohorts: One predictive study (cohort 1) and two intervention studies (cohorts 2 and 3). Method Participants completed a TPB measure at baseline and a measure of fruit and vegetable intake at 1 week (cohort 1; n = 90) or 1 month (cohorts 2 and 3; n = 296). Results Attitude moderated the impact of PBC on intention. PBC moderated the impact of intention on behaviour at 1 week but not 1 month. Conclusion The variance accounted for by the interactions was small. However, the presence of interactions between constructs within the TPB demonstrates a need to consider interactions between variables within the TPB in both theoretical and applied research using the model.