The Strengths and Difficulties Questionnaire (SDQ) – Self-Report. An analysis of its structure in a multiethnic urban adolescent sample Jo ¨rg Richter, 1 A ˚ se Sagatun, 1 Sonja Heyerdahl, 1 Brit Oppedal, 2 and Espen Røysamb 2,3 1 Centre for Child and Adolescent Mental Health Eastern and Southern Norway, Oslo, Norway; 2 Norwegian Institute of Public Health, Oslo, Norway; 3 Department of Psychology, University of Oslo, Oslo, Norway Background: The SDQ is currently one of the internationally most frequently used screening instru- ments for child and adolescent mental health purposes. However, its structure, cross-cultural equiv- alence, and its applicability in ethnic minority groups is still a matter of discussion. Methods: SDQ self-report data of 5,379 ethnic Norwegian and 865 ethnic minority adolescents with a variety of national origins was analysed by means of confirmatory factor analysis (CFA). Multi-group comparisons considering equal thresholds combined with more in-depth analyses on factor loadings, residuals, composite reliability, and average amount of variance explained by indicators of respective constructs were performed. Results: CFA suggested a good fit of the five-factor model of the SDQ self-report in the subsample of ethnic Norwegian adolescents and an acceptable fit in ethnic minority subsamples without substantial differences between ethnic Norwegian SDQ data and data of Pakistani or those of ‘other ethnic minority’ adolescents. When assuming equal thresholds between response categories of the items as well as equal factor loadings the structure in the data significantly differed between ethnic Norwegian and both ethnic minority samples. Some factor loadings and some correlations between constructs significantly differed between ethnic Norwegian and both ethnic minority samples. The correlation coefficients between the hyperactivity factor and the conduct problems factor were too high in all three subsamples in order to establish distinct constructs. Composite reliability and average explained variance of the emotional symptoms factor were good in all samples, whereas they were low for some of the other factors. Conclusions: To some extent the theoretically proposed five-factor structure of the Norwegian version of the SDQ self-report was supported in 15- to 16-year-old adoles- cents. However, the results of more detailed analyses raise questions about the interpretation of some subscales. When applying this screening method to Norwegian adolescents, our results suggest that the use of the total difficulty score of the SDQ in screening youth should be preferred over the subscale scores. Keywords: Self-report SDQ, factor structure, CFA, adolescents, ethnic minorities. The Strengths and Difficulties Questionnaire (SDQ; Goodman, 1997; Goodman, Meltzer, & Baily, 1998) is currently one of the most frequently used screening instruments for child and adolescent mental health throughout the world. The instrument includes four problem subscales of five items each that also yield a total difficulties score, in addition to a prosocial dimension (http://www.sdqinfo.org). Its psychomet- ric properties have been extensively investigated. The various SDQ versions (self-report, parent, teacher) have been translated into more than 60 languages, including languages from many non-Western coun- tries: 20 Asian and 6 African language versions were available by November 2010 (http://www.sdqinfo. com). Results from epidemiological studies have been published from many countries. A few cross-cultural studies have compared prevalence and level of dis- tress findings across countries (Obel et al., 2004; Mojtabai, 2006; Heiervang, Goodman, & Goodman, 2008). Psychometric properties have also been investigated in many countries with different language versions of the scale. Most of the studies of psychometrics have investigated properties of the SDQ when used in a specific country, but Goodman, Renfrew, and Mullick (2000) applied an algorithm developed in the UK to predict clinical diagnosis in data from Bangladesh. They found that the algorithm worked just as well in the Bangladeshi sample, demonstrating comparable predictive validity. Cross-cultural equivalence is essential when comparing results from various cultures and sub- groups. However, to the best of our knowledge no study has comprehensively analysed cross-cultural equivalence of the measure, i.e., of configural invariance, metric invariance, tau invariance and scalar equivalence, by CFAs. Some evidence was provided by a recently published article where mea- surement equivalence relating to factor loadings, thresholds between response categories and residu- als of the SDQ items was found to be the same for data from 5- to 16-year-old British Indians and Whites (Goodman, Patel, & Leon, 2010). Conflict of interest statement: No conflicts declared. Journal of Child Psychology and Psychiatry 52:9 (2011), pp 1002–1011 doi:10.1111/j.1469-7610.2011.02372.x Ó 2011 The Authors. Journal of Child Psychology and Psychiatry Ó 2011 Association for Child and Adolescent Mental Health. Published by Blackwell Publishing, 9600 Garsington Road, Oxford OX4 2DQ, UK and 350 Main St, Malden, MA 02148, USA