Validity of the Middle Years Development Instrument for population monitoring of student wellbeing in Australian school children

Abstract

The importance of social and emotional wellbeing has long been recognised by education systems but the measurement of wellbeing still receives far less attention than the measurement of academic achievement. This paper reports on a five-year project to measure student wellbeing across an education system within the state of South Australia using the Middle Years Development Instrument (MDI). All schools (Government, Catholic, and Independent) were invited to participate in the collection at no cost and aggregated school reports provided an incentive to participate. A total of 51,574 students completed the MDI between 2013 and 2015, with higher participation rates in Government schools than Catholic or Independent schools (65%, 18 and 13% respectively in 2015). Validity and reliability analyses confirmed that the MDI scales had good psychometric properties (i.e., favourable model fit in confirmatory factor analyses, high internal consistency, and correlations between scales were consistent with theoretical expectations). Test-retest reliability (based on a sub-sample of 82 children) was acceptable for most scales except for the connectedness to adults at school (r = .50) and friendship intimacy scales (r = .40), where test-retest reliability was low. However, several of the MDI scales had ceiling effects, particularly for girls and younger students (10–11 years old), which may present challenges when using these scales for population monitoring, program and policy evaluations. Pragmatic factors for education systems and governments to consider in selecting social and emotional wellbeing tools are discussed.