Teun Remmers

116 | Chapter 7 Strengths and weaknesses The major strength of this study is that we attempted to improve the understanding of ASPA and a potentially plausible relationship with playability of the physical environment, by measuring time-period specific ASPA and detailed, qualitative characteristics of playgrounds in school-environments. The present study was confined to an ecological design, assessing characteristics at the school level. Recent methodological innovations such as combined accelerometry and GPS measurements can provide opportunities for even more in-depth analyses of the association between environmental attributes and domain- specific PA. GPS measurements can for example be used to identify time spent outside (26) or even at the schoolyard or at specific playgrounds (42). Moreover, integrating multiple data-sources (e.g. accelerometry, GPS, GIS, audits, school’s time tables, participant diaries) into comprehensive databases provide unique opportunities for investigating PA and other health behaviors, while accounting for its spatial and temporal specificity (8, 43). This study applied a threshold of 50% period-specific registration time to prevent our analyses being influenced by short- spurious spikes of (intense) PA. Although two studies also used this 50% threshold (27, 34) and one study used a 60% threshold (26), reliability and relative influence of exact thresholds for period-specific registration times are debatable. In addition, the definition of non-weartime periods (e.g. 60 minutes of consecutive zero’s) highly influences period-specific thresholds, as it determines whether relatively short- or longer bouts of inactivity are classified as non-weartime periods. Future studies are therefore warranted to investigate the influence of period-specific thresholds in depth while also accounting for differences in non-weartime definitions; for example by comparing its influence in relationships with ASPA patterns. One can speculate on alternative protocols in computing playability (e.g. multiplication of individual items or based on unequal weights). However, as we were unable to find an evidence-base for such alternative protocols, we decided to aggregate based on equal weights. To check for potential errors in the aggregation procedure we fed back the aggregated scores to the auditors of school-environments. Hereafter, no alterations were made to the aggregation protocol. As our audits were limited to 800 meter buffers from participating schools, it may seem logical that the relationship between playability of the school-environment and ASPA attenuated for children that lived outside this study-area. However, the aim of the present study was to demonstrate the temporal and site-specific mechanisms, and thus underline that in future research investigating relationships between PA and the environment, time- and place specificity is warranted. In addition, as in the Netherlands no public primary schools have organized public transport services from school to children’s homes and the majority of 8-11 children rely on active transport to get home after school, all children are likely to have at least the opportunity to be exposed to PA-opportunities in their school- environment.