Teun Remmers

114 | Chapter 7 Discussion This study investigated the association between playability of school-environments and ASPA, separately for time-periods within the after-school period and distance-categories from school to children’s residence. We demonstrated that the influence of playability is highly dependent on these time-periods and distance from home to their school: greater distance attenuated the influence of playability of the school-environment on ASPA, especially in the first hours after-school. As expected, children that lived outside the study-area for which playability was audited generally showed no relationships between playability of the school-environment and ASPA. When comparing our findings in the light of other studies that have investigated ASPA objectively, we can confirm that boys were more active after-school compared to girls (17, 20, 25). In contrast to the study of Mota et al., which reported that boys were more active in the later time-periods after school (19), we found that the difference between boys and girls was stable across the first three time-segments, and this difference decreased at later periods in the evenings (data not presented). As this study did not compare PA during school hours with ASPA, we cannot compare our results with studies that indicated that ASPA significantly contributed to total PA (17, 20, 21). Although Timperio et al. found that relationships between ASPA and individual features of public open spaces were different for boys than for girls (22), we found no such moderation mechanisms in our playability index. This may be because in our study, potential gender differences may annul at a higher level of abstraction when utilizing a standardized index-score of playground qualities instead of individual features of public open spaces. Further research is however needed to clarify potential gender-related moderation mechanisms. In addition, Scott et al., argued that perceptions of easy access and the number of PA- facilities, but not objectively determined number and proximity of PA-facilities were related to adolescent girls’ non-school PA (24). Although we acknowledge the importance of perceived accessibility and/or presence of environmental attributes in PA-research, we cautiously suggest that this played a relatively minor role in our study because our playability index aggregated qualities of multiple playgrounds, accounting for accessibility and the number of these playgrounds in the school-environment. When taking into account the relatively small strata-specific sample sizes at greater distance-categories, relationships may seem relatively weak. However, this study demonstrated that relevant relationships between physical environments and ASPA can be revealed and made plausible, with increasing specificity in time and distance. This demonstrates that loss of statistical power due to lower number of observations is compensated by increased discriminative precision thanks to time-place specificity. Our proposed playability score allowed for aggregation of playground characteristics of the school-environment within multiple geographic settings. The SPACE observation instrument is comparable to the Neighborhood Environment Walkability Index in terms of identified factors/scales (e.g. facilities, aesthetics, proximity, accessibility), aggregation