*2.2. Procedure*

We e-mailed our online survey to 4716 (74%) of Denmark's approximately 6400 preschool institutions and BASCs. We selected the institutions randomly, adjusting only to represent both urban and rural areas in all parts of Denmark. BUPL provided the list of institutions. We supplemented this with institutions from municipalities in Denmark underrepresented on BUPL's list. We obtained the e-mail addresses of these institutions from the municipalities in which they operated. As we were unable to obtain teachers' e-mail addresses, we sent the invitations to participate to the directors, and invited them to respond to the questionnaire and to pass on the survey link to as many of their teachers as possible. Initially, substantially more directors than teachers answered our survey, and through the directors, we targeted the teachers with three reminders. This reduced the difference between the numbers of directors and teachers participating in our survey, although a considerable difference remained in the final result (1374 directors versus 677 teachers). The percentage of male survey participants was high (22%), relative to the percentage of male pedagogues at a national level (12%), which is important, as the concern under discussion is particularly sensitive for male childcare workers [21].

Our questionnaire had a 30.9% response rate, which represents 23% of all preschool institutions and BASCs in Denmark. The pilot study had a 20% response rate, and when we contacted the unresponsive institutions, they all explained that they were pressed for time and received many surveys. Guided by the pilot study, we made the questionnaire easier to answer, which, with the reminders, may account for the improved response rate of the final survey.
