Prediction of Behaviour vs. Prediction of Behaviour Change: The Role of Motivational Moderators in the Theory of Planned Behaviour

Research has identified moderating variables of the cognition–behaviour relationship in the Theory of Planned Behaviour (TPB). These include intention certainty, affective–cognitive congruence of attitudes, attitudinal ambivalence, and past behaviour. It has not yet been systematically tested if these moderators apply in similar ways to the prediction of behaviour and behaviour change (when past behaviour is controlled). A sample of 1,418 university students completed measures of the direct predictors of the TPB (perceived behavioural control; PBC and intention), intention certainty, affective–cognitive congruence of attitudes, attitudinal ambivalence, and the Godin Leisure Time Physical Exercise Questionnaire at the beginning of the academic year and two months later. Data were analysed using moderated hierarchical linear regressions. Without controlling for baseline behaviour, intention certainty and affective–cognitive congruence moderated both the intention–behaviour and PBC–behaviour relationships and showed additional main effects. Attitudinal ambivalence did not contribute to the prediction. When past behaviour was controlled, the main effects were attenuated and only the intention certainty* PBC interaction remained significant. Predictors of behaviour do not necessarily predict behaviour change. The TPB‐related moderators studied here weakly contributed to the prediction of behaviour change. Proposals for increasing the predictive power of TPB for behaviour change are discussed. Point de départ. Des recherches ont mis en évidence des variables régulatrices de la relation cognition‐comportement dans la Théorie du Comportement Planifié (TPB). Celles‐ci comprennent la solidité de l’intention, la conduite passée, l’ambivalence et la congruence affectivo‐cognitive des attitudes. On n’a pas encore systématiquement recherché si ces régulateurs s’appliquaient de la même façon à la prédiction du comportement et au changement comportemental (quand le comportement antérieur est contrôlé). Méthode. Au début de l’année universitaire et deux mois plus tard un échantillon de 1,418 étudiants a fourni des données relatives aux prédicteurs directs du TBP (le contrôle comportemental perçu, le contrôle du comportement passé et l’intention), à la solidité de l’intention, à l’ambivalence et à la congruence affectivo‐cognitive des attitudes et répondu au questionnaire de Godin portant sur l’exercice physique durant le temps de loisir. Les données ont été analysées avec