Page 124 - ON THE WAY TO HEALTHIER SCHOOL CANTEENS - Irma Evenhuis

Chapter 7. General Discussion
the lower educational levels). This makes the results less generalisable to younger and older age groups, as younger students have different (determinants of) dietary behaviour than older students. To illustrate, younger students might take food and drinks from home more often, might have less money to spend, and are likely to be less independent compared to older students [210]. Besides, as the influence of parents decreases and the influence of peers becomes more prominent, the number of purchases at and around school may differ between ages [19, 20]. Measuring the effect in only one age group might have made the effect of the healthier school environment less visible.
To evaluate the effect and process of the implementation plan, we performed a quasi- experimental controlled trial. This is a potential limitation because Randomised Control Trials (RCT’s) are generally recognised as the best design to investigate the effect of an intervention because random assignment assures no systematic difference between the intervention and control group. However, in Dissemination and Implementation (D&I) Research, due to the external validity, practical concerns, and the involved contextual factors, randomisation is not always feasible, making other designs more suitable [211]. In our study, due to the small number of included schools and the fact that different contextual school factors are related to implementation, randomisation was not appropriate. Meanwhile, it is a strength that we matched our intervention schools to control schools in order to reduce the pre-existing differences between intervention and control schools [212]. We matched schools on the characteristics educational level, school size, and catering by the school itself or external catering company. These were broadly similarly distributed across the matched characteristics. Although we were not able to match additionally on the availability of shops near school, and the presence of policy to oblige students to stay in the schoolyard during breaks, these were also about equally distributed across both groups. Moreover, gaining insight into other contextual factors and dynamic processes (in for example the school organisation) which also may influence implementation of school health policy, was part of our intervention [52, 68]. For example, by using the schools’ and stakeholders’ questionnaires and providing tailored support based on these insights.
What is the perfect timing in research?
In our studies, multiple strengths and limitations occurred with regard to the timing and duration, including the timing of the measurements. One strength is that the interests of both practice and research were balanced in the timing and study duration. We wanted to gain insight into how and when to properly implement the guidelines, while also wanting to start using the developed guidelines as soon as possible. The involvement of multiple stakeholders, like the school canteen advisors and caterers, meant that they had the opportunity to become familiar with the guidelines prior to broader dissemination during the research. Besides, their experiences with and research insights on the implementation tools enabled adaptations before widespread dissemination. This made waiting with broader dissemination of the guidelines until the research results were available worthwhile.
Unfortunately, we had a limited amount of time for the evaluation of the implementation plan, due to the development of the implementation tools on the one hand (like the moment the Canteen Scan was available), and the end of the school year on the
122