Abstract
Objectives. Interest in attitudes and ambivalence has highlighted problems with the adequacy of conceptualizing attitudes as unitary, unidimensional, evaluations. In this paper, we report an application of the Theory of Planned Behaviour (Ajzen, 1991) in the domain of dietary choice which investigates the hypothesis that ambivalence will attenuate observed attitude-intention relationships (since the evaluations influencing the expression of attitudes are more Likely to differ from the evaluations influencing the expression of intentions for people who are more ambivalent. Design. Participants complete a postal questionnaire which contained, inter alia, measures assessing the components cf the Theory of Planned Behaviour and a measure of ambivalence. Methods. Participants (N = 296) were recruited via advertisements placed in local newspapers, asking for volunteers to assist in a research project. Participants were randomly assigned to complete a questionnaire about either their chocolate consumption or their meat consumption. Participants were paid pound4. Results. The findings show considerable support for the hypothesis: there was a tendency for attitude-intention relationships to be attenuated among participants with higher levels of ambivalence, compared to participants with lower levels of ambivalence. Conclusions. The research supports the widespread view that ambivalence is an important issue, both for those involved in basic attitude research and for those who seek to use attitude theories in applied research. In particular, the findings indicate that ambivalence may often have implications for the predictive ability of attitude-intention-behaviour models, especially when applied to health-related behaviours that are characterized by motivational conflicts.