Self-affirmation and health behaviour change : affective and cognitive predictors, moderators and mediators.

This thesis explored the potential of self-affirmation for use as a health intervention by exploring the effect of self-affirmation on a range of health behaviours (e.g., salt, alcohol, and fruit and vegetable consumption). The moderating effects of risk, motivation (e.g., importance of health, decisional balance, prevention focus) and threat on the effects of self-affirmation were also explored in order to determine under which circumstances self-affirmation manipulations were most effective. The thesis also explored potential mediators of self-affirmation (e.g., explicit and implicit self-related affect and efficacy cognitions). In four experimental studies, participants completed a self- or non-affirming task prior to exposure to a health message. The effects of self-affirmation on cognitive, affective and behavioural responses were examined. The studies found that self-affirmation increased a range of cognitive (e.g., self- and response-efficacy) and affective predictors of health behaviour change (e.g., negative self-evaluative affect, implicit threat-related affect) and actual behaviour change (e.g., consumption of fruit and vegetables). The thesis found that response-efficacy mediated the effect of self-affirmation on behaviour change. The thesis also found that there was a trend for self-affirmation manipulations to be most effective with lower threat health messages, and on participants who were most vulnerable (i.e., at the highest risk or the lowest motivation). However, the studies also suggested that self-affirmation manipulations may be detrimental when high threat health messages are used or on low risk or highly motivated participants. The implications of the studies are that self-affirmation manipulations are capable of changing health behaviour; this offered preliminary support to the argument that self-affirmation manipulations have the potential to be developed as health interventions. However, the thesis also suggested that self-affirmation interventions would need to include low threat health information and that careful targeting would be needed when delivering the intervention.