Validiteten af spørgeskemaundersøgelser med lav svarprocent

Solvejg Kristensen, Charlotte Maria Jensen, Trine Nøhr Winding, Niels Bilenberg

Publikation: Bidrag til tidsskriftTidsskriftartikelForskningpeer review


INTRODUCTION: The purpose of the present study was to compare the characteristics of epidemiological survey non-respondents with the characteristics of respondents who participated by mail and phone. MATERIAL AND METHODS: The study was based on data from questionnaires and registered data on socio-demographics. The Child Behavior Checklist for Ages 1-5 (CBCL/1-5) and "Experiences when answering a questionnaire" was mailed to 300 parents of children aged 18 months to five years. Parents who did not respond after a mail reminder were contacted by telephone. RESULTS: A total of 138 families replied to the mail contact, 99 replied after telephone contact, and we received no response from 63 families. No statistical differences in socio-demographic characteristics were found between the three groups. No statistical differences were found concerning the total problem behaviour score of the CBCL/1-5 between mail and phone responders. Parents who participated by phone thought more frequently than mail responders that questions on the CBCL/1-5 were ethically offensive, and that their own knowledge about child well-being and behaviour was unsatisfactory, p < 0,01. CONCLUSION: Non-participation is associated with parental attitudes towards ethical issues related to CBCL/1-5 and to poor parental knowledge about child well-being and behaviour. In spite of this, and although there was a low initial participation rate, it seems reasonable to assume that the results of child psychiatric epidemiological questionnaire surveys can be generalised.
Udgivelsesdato: 2009-Feb-2
Bidragets oversatte titelValidity of epidemiological surveys with low participation rates
TidsskriftUgeskrift for læger
Udgave nummer6
Sider (fra-til)415-9
Antal sider4
StatusUdgivet - 2. feb. 2009