56 | Chapter 12 never having taken part in a sport during their lifetime, or never having done so regularly or intensively once in a while, as no life course data were gathered on these individuals. The final dataset therefore comprised 2707 respondents who during their lifetime had participated in at least one sport regularly or did so intensively once in a while. For each respondent, we created person-year combinations for all relevant years, extending from the 18th year of life up to and including the 65th. The major life events studied took place largely within this segment of life. Note that, in the Netherlands in 2010, everyone 65 years of age and older had the right to stop paid employment. However, a large share of the population retired earlier. In 2010, 73.5% of all employees retired before turning 65 years of age, and the average retirement age was 62.8 (Statistics Netherlands, n.d.). For all respondents, we studied the life course starting from age 18; therefore, the person-year combinations were created starting in the 18th year of life. For respondents who were older than 65 in 2010, no person-year combinations were created, as none of the six life events studied were experienced after the 65th year of life. Respondents who were younger than 65 in 2010 were included in the person-year file from their 18th up to and including their age in 2010. Respondents were “at risk” of starting a sport in all the years of the personyear file. The risk set thus comprised all of the person-year combinations for the respondents in the person-year file. At this point, it must be noted that respondents could report on a maximum of ten sports; so only the ten sports they participated in most often were reported. Six respondents from the Sporters Monitor 2010 said they had already participated in ten sports before their 18th birthday. Therefore, in terms of the risk set, they were formally no longer “at risk” of starting a sport after their 18th birthday. These respondents were therefore omitted. Similarly, the four respondents with 25 or more person-year combinations in which they had already taken part in ten sports, and thus were no longer at risk of starting a sport, were eliminated from the risk set. This left 532 person-year combinations in which a respondent had participated in ten sports. This was just 0.8% of our total number of 70,631 person-year combinations. We therefore chose not to remove these from the risk set. On average, each respondent had 26.1 person-year combinations in the risk set (70,631/2707).
RkJQdWJsaXNoZXIy MjY0ODMw