Do social cognitive constructs predict physical activity behaviour within individuals? a series of N-of-1 studies

Objectives
Previous research supports the ability of Social Cognitive Theory (SCT) in explaining physical activity (PA) behaviour. However, further tests of this theory
are required at the within-person level, because while SCT proposes that its constructs should explain behaviour within individuals, most studies have tested this theory between individuals in large group studies. The aim of this study is to examine whether SCT constructs can predict PA within individuals of varying activity levels.

Design
Six correlational n-of-1 studies were conducted.

Methods
Six adults aged 29-65 with varying levels of PA (e.g. two inactive, two minimally active, and two currently active) completed questionnaires on random days over a four-week period measuring SCT cognitions (e.g. barrier self-efficacy, goal setting, planning,
social support, outcome expectations, perceived barriers, enjoyment). Daily PA was assessed using a combination of accelerometers (step counts) and self-report log-books. Data were analysed using cross-correlational time series analysis.

Results
Cross-correlation analysis showed that at least one SCT construct predicted PA in five participants, although no individual had the same pattern of predictors across the study. Barrier self-efficacy and perceived barriers predicted behaviour in individuals
with lower levels of PA, but currently active individuals did not display similar patterns of
predictors.

Conclusions
The ability of SCT to predict PA within individuals was partially supported. Interventions aimed at reducing perceived barriers and improving self-efficacy may increase levels of PA in low active individuals. Tailored behaviour change techniques that target cognitions at the individual level are recommended.