Normative data and psychometric properties of the strengths and difficulties questionnaire among Japanese school-aged children

Abstract

BackgroundAlthough child mental health problems are among the most important worldwide issues, development of culturally acceptable mental health services to serve the clinical needs of children and their families is especially lacking in regions outside Europe and North America. The Strengths and Difficulties Questionnaire (SDQ), which was developed in the United Kingdom and is now one of the most widely used measurement tools for screening child psychiatric symptoms, has been translated into Japanese, but culturally calibrated norms for Japanese schoolchildren have yet to be established. To this end, we examined the applicability of the Japanese versions of the parent and teacher SDQs by establishing norms and extending validation of its psychometric properties to a large nationwide sample, as well as to a smaller clinical sample.MethodsThe Japanese versions of the SDQ were completed by parents and teachers of schoolchildren aged 7 to 15 years attending mainstream classes in primary or secondary schools in Japan. Data were analyzed to describe the population distribution and gender/age effects by informant, cut-off scores according to banding, factor structure, cross-scale correlations, and internal consistency for 24,519 parent ratings and 7,977 teacher ratings from a large nationwide sample. Inter-rater and test-retest reliabilities and convergent and divergent validities were confirmed for a smaller validation sample (total n = 128) consisting of a clinical sample with any mental disorder and community children without any diagnoses.ResultsMeans, standard deviations, and banding of normative data for this Japanese child population were obtained. Gender/age effects were significant for both parent and teacher ratings. The original five-factor structure was replicated, and strong cross-scale correlations and internal reliability were shown across all SDQ subscales for this population. Inter-rater agreement was satisfactory, test-retest reliability was excellent, and convergent and divergent validities were satisfactory for the validation sample, with some differences between informants.ConclusionsThis study provides evidence that the Japanese version of the SDQ is a useful instrument for parents and teachers as well as for research purposes. Our findings also emphasize the importance of establishing culturally calibrated norms and boundaries for the instrument’s use.

Caretaker mental health and family environment factors are associated with adolescent psychiatric problems in a Vietnamese sample

“…Internal reliability for the SDQ subscales varied from good to very low reliability: emotional symptoms (α = .70), conduct problems (α = .53), hyperactivity/inattention (α = .47), peer relationship problems (α = .26), and prosocial behavior (α = .61). The range of internal reliability estimates is broadly consistent with previous work on the scale (e.g., Colins et al, 2013), and the particularly low internal reliability for the conduct problems and peer relationship problems is consistent with youth samples in Japan (Moriwaki and Kamio, 2014), China (Mellor et al, 2011), and multicultural samples in the Netherlands (Colins et al, 2013; Mieloo et al, 2014). Past work suggests that differences in cultural reports of youth behavior may influence caretaker reports, particular in cross-cultural or minority group samples (Mieloo et al, 2014), thereby influencing reliability estimates.…”

Child prosociality within HIV-affected contexts: the impact of carer ill-health and orphan status

“…Respondent scores for the Prosocial Scale were also compared with clinical cut-offs 2 derived from British samples: average 6-10 (80% sample); borderline 5 (20%); and clinical 0-4 (10%) (Goodman, 1997). Child self-reports for the Prosocial Scale in this study (ɑ=.89), yielded stronger reliabilities than previous investigations (.69-.82) (Deighton et al, 2014), while results for the carer report (ɑ=.69) were within an accepted range (.63-.85) (Deighton et al, 2014). …”

“…However, limitations are noted in the use of a single (5-item) instrument, despite the adequate psychometric properties in this sample. The three-point Likert scale design limits variability in data, and may not be sensitive to detect changes over time (Deighton et al, 2014). Furthermore, cross-sectional studies do not provide a longitudinal perspective on psychosocial adjustment (Sherr at al., 2014).…”

Associated factors of behavioural problems in children at preschool age: the Hokkaido study on environment and children's health

“…; Matsuishi et al . ; Moriwaki & Kamio ). Previous study of 5 to 10‐year‐old children in UK and of 4 to 6‐year‐old in Japan showed similar total difficulties score to ours, 8.6 and 8.27, respectively (Matsuishi et al .…”