An Examination of Threshold Setting in Social Emotional Measurement

Abstract

Background and PurposeThresholds are used by nurses in social emotional screening. A purpose of this study is to examine fluctuation of the threshold pattern in a widely-used instrument, Ages and Stages Questionnaire-Social Emotional, Second Edition (ASQ:SE-2).MethodsThreshold settings are investigated against an irreversible pattern of child growth from research literature. Empirical studies are conducted on an issue of 6,039 missing cases in the ASQ:SE-2 normative sample that undermines data quality for the cutoff score configuration.ResultsA Bayesian estimate has been suggested to improve the threshold at age 3 with an asymptotically unbiased cutoff score. Given the coverage of ASQ:SE-2 for children from less than 6 months to 72 months for mental health referrals, this study is particularly relevant to the practice of social emotional screening performed by pediatric nurses, school nurses, and nurse practitioners at mental health hospitals.ConclusionTo address the violation of Classical Test Theory and Item Response Theory in the ASQ:SE-2 threshold settings, consistent statistical imputations are needed to maintain monotonicity of the cutoff score patterns that are aligned with the irreversible trend of child growth, as well as the well-established national standards for missing data examination.