In the absence of longitudinal data, recall data is used to examine participation in sport. Techniques of survival analysis are adapted and applied to illuminate the dynamics of sporting life. The likelihood of participation has a distinct pattern across the life-course, rising to a peak at 15 years of age, falling sharply in late teenage years and more gradually during adulthood. Logistic regressions and Cox regressions reveal strong effects on participation of gender, cohort and socioeconomic status, which vary over the life-course and by type of sport. The findings add significantly to previous work and have implications for policymakers wishing to increase physical activity.
sporting participation health survival analysis recall data