The university student health and wellbeing survey: test-retest reliability within a large Australian university undergraduate student population

Abstract

Issue addressed: Understanding the health behaviours of Australian university students and their impact on wellbeing and academic success is important, however there is no reliable, population level survey tool to measure student health across a range of domains. The purpose of the research was to determine the reliability of the web-based University Student Health and Wellbeing Study (USHWS) survey via a test-retest reliability study within a large Australian university student population. Methods: A test-retest of the measurement instrument was completed by a sample of university students (n=195) over a two-week period. The instrument assessed alcohol and tobacco use, mental health, sexual health, physical activity, nutrition and sun protective behaviours. Test-retest reliability was analysed using two-way random effects model of intraclass correlations (ICC) and AC1 coefficient for individual measurements with 95% confidence intervals (95% CI). Results: Questions on demographics, general health, nutrition, sun protection, alcohol and tobacco use, mental health and sexual health had fair to high reliability (ICCs range from 0.32-1.00). Reliability of some physical activity items were poor with large variability (ICC = 0.15, 95% CI 0.01-0.28 to 0.86, 95% CI 0.82-0.89). Conclusions: A majority of the USHWS survey items represented a moderate to high test-retest reliability. Variability and poor reliability of physical activity questions may be due to survey implementation time and usual behaviour changes. So what? The USHWS survey is reliable instrument to assess Australian university student health at a population level with the aim of informing effective programming, policy and initiatives. SUMMARY The University Student Health and Wellbeing Study (USHWS) survey is a foundational tool to understand university student’s health in Australia. The USHWS reported fair to high reliability with few physical activity items showing lower reliability. Greater variability may be due to usual day-to-day fluctuations in behaviour.