Utility of brief teacher rating scales to identify children with educational problems

Abstract

The study aimed to determine the utility of teacher ratings of children's behaviour, oral language, and literacy skills in identifying children with educational problems, and to examine the relationship between behaviour, oral language, and literacy skills. Teacher ratings were obtained for a cohort of Year 2 pupils (N = 129, mean age = 7.3 years, SD =.33) from five schools in Western Australia, using three brief screening questionnaires. To determine the sensitivity and specificity of the screening instruments in identifying children with educational problems as determined by psychometric criteria, direct psychometric assessment of oral language, vocabulary, reading, and spelling was conducted for a subset of the children. Teacher ratings of language, literacy, and behaviour correlated significantly with the standardised test scores. The screening instruments for language and literacy had reasonable sensitivity (74%) and specificity (92%), but a fairly high rate of false negatives (26%) and relatively poor positive predictive power (.53) for identifying children with problems in language and/or literacy. The prevalence of marked behaviour problems in this Australian sample of school children (9.6%) was comparable to that found in previous studies. Brief teacher-based screening instruments appear to have the potential to provide an economical and effective approach for identifying children at risk for educational problems, but the current instruments lacked adequate precision. The marked overlap between educational and behavioural problems in young children has important implications for prevention, identification, and intervention policies.