Predicting parental social influences: The role of physical activity variability

Objectives: This study examined how variations in the activity of adolescents might relate to the social influences used by parents. Specifically, mean differences in activity across adolescents (individual differences) as well as variation in activity within adolescents (intra-individual variation) were used to predict the use of family social influence. Design: A prospective design was used. Methods: High school students (N = 329) completed measures of activity (six times) and social influences received from family (positive, negative, collaborative) (five times) over a one-year period. A multilevel analysis was used to predict each of the three types of social influence. The predictors included activity in two forms: The individual's mean level of activity (i.e., individual differences) and activity at the previous time point as a deviation from the individual's mean activity level (i.e., intra-individual variation). Results: Controlling for age, gender and school, results revealed that both individual differences in mean activity level (b = 0.04, p < .001) and intra-individual variation in activity (b = 0.02, p = .055) predicted use of collaborative influence. For positive influence, only individual differences in mean activity level was a predictor (b = 0.02, p < .001). Neither form of activity predicted parental use of negative influence. Conclusions: Results revealed support for the positive relationship between both positive and collaborative types of influence and individual differences in activity that is typically reported in the literature. However, a negative relationship between intra-individual variation in activity and collaborative social influences also was found. This negative relationship, where lower activity than normal by the adolescent was related to a greater use of collaborative influence by the parent, may be indicative of a regulatory form of social influence. (C) 2011 Elsevier Ltd. All rights reserved.