The TPB has been used to assess a variety of health-related behaviours. For example, Brubaker and Wickersham (1990) examined the role of the theory’s different components in predicting testicular self-examination and reported that attitude towards the behaviour, subjective norm and behavioural control (measured as self-efficacy) correlated with the intention to perform the behaviour. A further study evaluated the TPB in relation to weight loss (Schifter and Ajzen 1985). The results showed that weight loss was predicted by the components of the model; in particular, goal attainment (weight loss) was linked to perceived behavioural control. Similarly, Conner, Lawton et al. (2006) used the TPB to predict speeding behaviour using a driving simulator and an on-road speed camera and showed a significant role for most of the TPB variables in predicting both intentions and actual behaviour. Recently, O’Connor et al. (2006) also used the TPB to predict deliberate self-harm and suicidality at three months’ follow-up as a means to explore whether the TPB was relevant to more extreme behaviour and whether social cognitive variables were better predictors than clinical variables. The results showed a strong role for variables such as self-efficacy, attitude and descriptive norm and that these were better predictors than depression. There have now been several reviews and meta-analyses of the TPB which describe the extent to which this model can predict a range of health behaviours (Sheeran and Taylor 1999; Armitage and Conner 2001; Trafimow et al. 2002).