15581-m-nanninga

SOCIAL SUPPORT AND PARENTING SKILLS 91 Differences between respondents and non-respondents were trivial to small regarding age, gender, degree of urbanisation and psychosocial problems, with Cohen’s effect sizes ranging from 0.01 (age) to 0.12 (degree of urbanisation) for the care sample. For this study only parent/caregiver-reported information was used, which was available for 1,331 children. The community sample ( N =1,025) concerned a stratified random sample of schoolchildren, obtained via five primary schools, two secondary schools and one school for intermediate vocational education from the same region recruited by taking into account the distribution of children across the study region according to their age, gender, socio-economic position and degree of urbanisation. Of these, 77 were excluded. Of the eligible 948 respondents, 99 filled out the opting-out form and 666 participated (70.3%). Differences between respondents and non-respondents were trivial for age, gender, degree of urbanisation and psychosocial problems, with effect sizes ranging from 0.02 (psychosocial problems) to 0.08 (degree of urbanisation). Parent-reported information was available for 664 children. Differences between respondents and children in the community were trivial for age and gender, effect sizes were 0.00 and 0.01, respectively. We limited the analyses to those participants in the community sample whose child had not had contact with a professional for psychosocial problems in the past 6 months, resulting in 463 participants. Data were obtained from parents/caregivers via web-based or paper questionnaires. Research comparing web-based and paper questionnaires demonstrates that there are no differences in the contents of responses and neither in the type of non-response [37,38]. If required, we provided assistance in filling out the questionnaire. Participants filled out the first questionnaire when they enroled in the study; for participants in the care sample this meant at the moment of their child’s care enrolment. They were frequently reminded about filling out the questionnaire, and returned questionnaires were checked for completeness to reduce the chance of missing data. Participants were rewarded with a gift token. Measures Psychosocial care enrolment was defined as enrolment of the child in any new type of care in one of the psychosocial care organizations because of psychosocial problems. These organizations were preventive child health care, child and adolescent social care, and child and adolescent mental health care. Family social support was measured with a subscale of the Family Questionnaire (FQ) [39], based on family circumstances that influence a child’s social-emotional development. The subscale measures support from relatives, friends, and neighbours,