Method: The complex and nested nature of this data requires the use of multi-level techniques so the session will describe results using Latent Growth Curves (a form of multi-level modeling for time sequenced data). The longitudinal patterns are first examined over time (level 1 predictor). “Student” related demographic variables are then added (level-2 predictors) followed by “school” related variables such as participation in NFL Play 60 initiatives (level-3 predictors) to determine the effect on the trajectories.
Analysis/Results: The models demonstrate modest impacts of programming on school-level physical activity and physical fitness scores over time. The focus of the project is on the extent to which programming impacts these outcomes so we specific attention was given to cross-level interactions between school variables and assessment periods defined at each year (“time” variable). The changes in activity and fitness show small effects when participation in NFL Play60 programming is directly examined. However, the findings are noteworthy considering the scope of the study and the participatory nature of the programming.
Conclusions: The results show the unique insights that are possible through this type of participatory research model. The continued exploration of this model will help to determine factors needed for successful adoption of Fitnessgram and school level programming in schools. Future applications of this work may help advance school programming focused on promotion of physical activity and physical fitness in youth.
See more of: Research