Multilevel modeling of walking behavior: advances in understanding the interactions of people, place, and time.

It has become increasingly clear that the influences on walking as well as other forms of regular physical activity are complex and require an increased understanding of factors across multiple levels of influence. Ecological frameworks have provided the field with a heuristic means of capturing potential impacts on behavior across diverse domains, including personal, behavioral, social or cultural, and environmental. We discuss advances in both understanding and applying this framework through the inclusion of previously ignored dimensions of impact (e.g., time), the application of state-of-the-art statistical methods for understanding interactions among multiple domains (e.g., signal detection), and the development of computer technologies (e.g., agent-based modeling) aimed at simulating the complex relationships between multiple levels of impact and walking behavior. We conclude with suggestions for future research in this emerging field.