Abstract
This article reviews recent developments in the design of interventions to improve health behavior. Based on dual-system models we classify intervention strategies according to whether they aim at: (i) changing impulsive structures; (ii) improving the ability to self-control; or (iii) changing reflective structures. We review recent work on re-training of automatic associations, attentional biases, and automatic approach–avoidance tendencies, training of self-control and executive functioning, and taxonomic work on health behavior intervention techniques. The theoretical framework as well as the empirical evidence suggest that a combination of both established and newly developed intervention techniques may prove fruitful for future intervention programs. However, several techniques are still in their infancy and more research is needed before clear recommendations can be given.
Acknowledgments
This work was supported by a grant for a research visit from the Swiss National Science Foundation to Malte Friese and a grant from the German Science Foundation (DFG) to Wilhelm Hofmann (Hofm 4175/3–1).
We thank Denise van Deursen for helpful comments on an earlier version of this manuscript.
Notes
Although the horse is not meant to be a direct expression of Bargh's (Citation1994) four horsemen of automaticity, impulsive processes are often less effortful, intentional, and controllable than reflective processes symbolized by the rider.
Beyond dual-system models, our framework is largely compatible with other approaches such as temporal self-regulation theory (TST; Hall & Fong, Citation2007) that also stresses the moderating role of self-control abilities for the link between reflective processes and behavior. One difference is that TST conceptualizes prepotent responses such as impulses also as such a moderator. Our framework understands impulses as a precursor of behavior the influence of which is moderated by—among others—the ability to self-control. In a series of studies, we did not find the reflective × impulsive interactions predicted by TST (e.g., Hofmann et al., Citation2008). Future research will identify the strengths and shortcomings of both frameworks.