A cusp catastrophe analysis of changes to adolescent smoking behaviour in response to smoking prevention programs

Abstract

The efficacy of smoking prevention programs aimed at adolescent smoking behaviour is widely debated in the health psychology literature. In general, however, these are not seen to be particularly effective in eliminating this acknowledged health risk behaviour. Even when positive results are presented, they frequently assume a linear association between exposure to some prevention or other and the dynamics of subsequent smoking behaviour change. Clair (1998) demonstrated that for alcohol consumption behaviour in adolescents, at least, this was not necessarily so. A nonlinear model, and in Clair's particular case, a Cusp Catastrophe Model (CCM) provided a better fit for the data than did any of a number of simple or interactive linear models. The present paper reports the use of precisely the same analysis for change in adolescent smoking behaviour following exposure to one or other of three smoking prevention programs of different orientations. While changes to adolescent smoking behaviour were evident following intervention, the reported analyses suggested that unlike alcohol consumption behaviour, CCM's were not necessarily the best nonlinear representation of the data.