Increasing evidence suggests that various health behaviours are amenable to change following the induction of cognitive dissonance. This systematic review sought to evaluate the effectiveness and methodological quality of dissonance-based health behaviour interventions and to explore identified sources of heterogeneity in intervention effects. Bibliographic databases were searched for relevant articles from inception to March 2012. Only studies targeting non-clinical health behaviour in non-clinical populations were included in the review. One author extracted data and assessed quality of evidence and a second author verified all content. Reports of 20 studies were included. A variety of health behaviours and outcome measures were addressed across studies. Most studies produced one or more significant effects on measures of behaviour, attitude or intention. Across studies, methodological risk for bias was frequently high, particularly for selection bias. Gender and self-esteem were identified as potential moderator variables. The evidence for the effectiveness of dissonance-based interventions was generally positive. The hypocrisy paradigm was found to be the most commonly applied research paradigm and was most effective at inciting change across a range of health behaviours. There was no observable link between type of target behaviour and positive outcomes. Researchers are encouraged to minimize potential for bias in future studies and explore moderators of the dissonance effect. What is already known on this subject? A recent meta-analysis indicates that dissonance-based interventions primarily based on the induced compliance paradigm are effective for eating disorder prevention (Stice, Shaw, Becker, & Rohde, 2008, Prev. Sci., 9, 114). However, it is currently unclear whether such outcomes are generalizable to interventions targeting non-clinical health behaviours such as smoking, sun protection and sexual risk taking. Other research indicates that studies based on the hypocrisy paradigm may lead to changes in non-clinical health behaviours (Stone & Fernandez, 2008, Soc. Personal. Psychol. Compass, 2, 1024; Stone & Focella, 2011, Self Identity, 10, 295) although this literature lacks systematic evaluation of interventions across a range of experimental paradigms. What does this study add? The hypocrisy paradigm appears most effective in inciting change across a range of non-clinical health behaviours. The dissonance effect may be moderated by variables such as self-esteem and gender. Risk of bias needs to be minimised to increase the validity of studies within this topic area.
Read full abstract