60 | Chapter 12 overview of the descriptive statistics for the dependent variables, the major life events and the control variables. Analytical strategy For the analyses of starting a sport (0/1) we used a logistic event history analysis. Our dependent variable starting a competitive club sport is comprised of five categories. We therefore made use of a multinomial logistic event history analysis, estimating the parameters based on the odds ratios of starting a competitive club sport versus every other category (Lammers et al., 2007). The parameters presented are the logit function (b), the standard error of the logit function (s.e.) and the odds ratio (Exp(B)). The logit function indicates the increase or decrease of the logit (the logarithm of the odds) per unit of an independent variable. The odds ratio is the multiplication factor of the odds and signifies the increase or decrease in the odds per unit of an independent variable. Essentially, a positive (negative) logit function and Exp(B) above (below) 1 indicate that the odds of starting a sport in general or in a competitive club setting increases (decreases) if the value of an independent variable increases, for example when a life event is experienced. RESULTS Starting a sport Table 2.2 presents the results of the logistic event history analysis of starting a sport. Model 1 incorporates the control variables: sex, age, period and migration background. The model shows that the odds of starting a sport were 11.3% smaller for men than for women (Exp(B)=0.887). Age was negatively associated with starting a sport (Exp(B)=0.968). Every year that a person aged, their odds of starting a sport decreased by 3.2%. The increased popularity of sport participation over time is evident from the effect of period (Exp(B)=1.030).
RkJQdWJsaXNoZXIy MjY0ODMw