Investigating the role of self‐control beliefs in predicting exercise behaviour: A longitudinal study

Abstract

Background Engaging in exercise behaviour regularly requires a repeated investment of resources to reap the health benefits. An individual’s self‐control resources, when performing a behaviour can be perceived as being recharged or depleted. The investigation on how self‐control beliefs resources predict exercise behaviour is very limited in the literature. The purpose of this study was to understand how self‐control beliefs predict exercise behaviour across time in a physical activity model. Methods Participants (N = 161) were a sample of adults recruited across twelve gyms and recreation centres in a large city. Participants completed surveys across five months. Data were analysed using a multilevel structural equation model with participants (level 2) nested within time (level 1). Results Behaviour was found to be a function of intention, habit, and planning. Specifically, planning moderated the intention‐behaviour relationship, where those who scored higher on planning engaged in more exercise. Self‐control beliefs functioned as a proximal predictor of autonomous motivation and predicted habit, and intention when accounting for total effects. Conclusions Self‐control beliefs played a pivotal role in supporting recognized exercise determinants. Exercise‐focussed interventions that help participants strengthen their beliefs as recharging and reduce depletion beliefs could be beneficial for promoting regular exercise.

Untitled

“…Related to this, autonomously motivated individuals are also more likely to form habits for those behaviors because they are likely to perform those behaviors in a consistent fashion with high frequency and in stable contexts-the key conditions under which habits form [34,35]. This has been supported by previous research on the autonomous motivation-habit relationship [36]. Based on these premises, it follows that autonomous motivation toward physical activity in PE and leisure-time contexts is expected to be indirectly related to intentions and physical activity behavior mediated by habits.…”

“…While self-control has been identified as an independent determinant of intentions and behavior, there is also research that has linked self-control, and self-regulatory processes in general, with the forms of motivation implicated in the motivational sequence of the transcontextual model. For example, research has highlighted that individuals reporting self-determined motives are less likely to be vulnerable to self-control failure and ego-depletion [40][41][42], and more likely to report intentions toward, and participate in, future behaviors, including physical activity [36,39,43]. These findings are consistent with the self-determination theory hypothesis that autonomous motivation is 'energizing' and individuals with autonomous motives toward behaviors are likely to report greater capacity to perform the behavior, and hence greater self-control [36].…”

“…For example, research has highlighted that individuals reporting self-determined motives are less likely to be vulnerable to self-control failure and ego-depletion [40][41][42], and more likely to report intentions toward, and participate in, future behaviors, including physical activity [36,39,43]. These findings are consistent with the self-determination theory hypothesis that autonomous motivation is 'energizing' and individuals with autonomous motives toward behaviors are likely to report greater capacity to perform the behavior, and hence greater self-control [36]. This hypothesis also aligns with the premise of the trans-contextual model that autonomous motives lead individuals to mobilize their resources to perform need-satisfying behaviors in future.…”

“…We also applied the Bayesian analytic approach to test model effects using informative prior values for key model effects derived from a meta-analysis of the model [18]. We also capitalized on previous research on self-reported habit [36,46] and trait self-control [38] to specify informative priors for these parameters in our test of extensions to the model. We expected to see a reduced level of uncertainty in the distributions of the parameters of the model specified for the current data when informative priors for key model parameters derived from the meta-analysis are specified, reflected in narrowed credibility intervals, compared to the distributions when non-informative priors are specified.…”

“…The defaults assumed a normal distribution with a mean of zero and a variance value essentially equivalent to infinity (10 10 ). In our second model, we applied informative prior values taken from previous research [18,36,38,49] to estimate model parameters. Specifically, priors for relations among the trans-contextual model constructs were taken from Hagger and Chatzisarantis' [18] meta-analysis.…”