The European Journal of Public Health Advance Access originally published online on February 5, 2008
The European Journal of Public Health 2008 18(3):294-299; doi:10.1093/eurpub/ckn001
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Infant, Child and Adolescent Health |
An international scoring system for self-reported health complaints in adolescents
Ulrike Ravens-Sieberer1, Michael Erhart1, Torbjorn Torsheim2, Jorn Hetland3, John Freeman4, Mia Danielson5, Christiane Thomas1 and The HBSC Positive Health Group
1 School of Public Health, WHO Collaborating Center for Child and Adolescent Health Promotion; University of Bielefeld, Germany
2 University of Bergen, Norway
3 Research Centre for Health Promotion, Bergen, Norway
4 Queen's University, Kingston, Ontario, Canada
5 National Institute of Public Health, Stockholm, Sweden
Correspondence: Prof. Dr Ulrike Ravens-Sieberer, School of Public Health, Director WHO Collaborating Center for Child and Adolescent Health Promotion, University of Bielefeld, PB 10 01 31, D-33501 Bielefeld, Germany, tel: +49 521 106 6245, fax: +49 521 106 6433, e-mail: U.Ravens-Sieberer{at}uni-bielefeld.de
Received May 31, 2007 , accepted December 24, 2007
Background: Aimed to develop a unitary scoring system for the ‘Health Behaviour in school-aged Children’ (HBSC) symptom checklist that would facilitate cross-national comparisons and interpretation. Rasch measurement analysis and investigation of differential item functioning (DIF) were conducted. Methods: Data were obtained from the ‘WHO collaborative study HBSC 2001/2002’. A total of 162 305 students aged 11, 13 and 15 years from 35 European and North American Countries were surveyed. Unidimensionality of the items and local independence were tested using means of confirmatory factor analysis. DIF across countries, age groups and gender was investigated using a logistic regression procedure. Item and person parameters were estimated according to the Rating Scale Model (RSM). Results: All items proved to be unidimensional. One item displayed noticeable DIF across countries and was discarded. The remaining items were functioning equally across subgroups. The RSM analysis resulted in Rasch model conform item parameter estimation. Infit mean square values between 0.84 and 1.35 revealed acceptable item fit. Conclusion: The control of DIF enables comparable and unbiased assessment of subjective health complaints across countries, age groups and gender. A scoring algorithm could be developed which enables a cross-cultural comparable and interval-scaled assessment of subjective health complaints.
Keywords: children and adolescents, differential item functioning, HBSC-Health Survey, international comparison, Rasch analysis, subjective health complaints