Exploring the influence of goals at different levels of abstraction on self-reported and electronically measured exercise frequency: an experimental field study

Although regular physical activity is associated with numerous health benefits, many people are not sufficiently active. Interventions that aim to increase physical activity rely mainly on concrete, “subordinate” goals. Based on a goal-theoretical perspective, we argue that combining goals at different levels of abstraction may foster successful goal pursuit, particularly in the long run. In the present study, all participants committed to the subordinate goal of exercising three times per week for three weeks. We used a 2 × 2 between-subjects design to assign participants to an additional superordinate goal, concrete action steps, or both; a control group focused solely on the subordinate goal. The main outcome was exercise frequency, which was measured (a) in the short term, i.e., during the three-week intervention period, using self-reports and electronic data; and (b) in the long term, i.e., during a six-month follow-up period, using electronic data. For the self-reported frequency in the short term, the results show an interaction between a superordinate goal and action steps: In the absence of action steps, a superordinate goal had a negative effect, but this negative effect dissolved when action steps were present. Similarly, action steps exerted a positive effect in the presence of a superordinate goal, but this effect dissolved in the absence of a superordinate goal. Goal manipulation had no significant influence either in the short or long term for electronically measured exercise frequency. Possible explanations for the observed effects and the differences between self-reported and electronically measured exercise frequencies are discussed.