Professor Joy Cumming is Director of the Assessment, Evaluation and Student Learning research area in LSIA at ACU. Joy has been involved in educational research for 40 years, including state and federally funded research projects and contributing to national and state policies. Her core research priority is educational assessment, including equity in assessment for students with diverse backgrounds. Her work has included an education law perspective examining the impact of educational policy and legislation in assessment and accountability on students.
Professor Harvey Goldstein is a Professorial Research Fellow with the LSIA, ACU. He is a chartered statistician, is joint editor of the Royal Statistical Society’s Journal, Series A, has been a member of the Society’s Council and was awarded the Society’s Guy medal on silver in 1998. Professor Goldstein is noted internationally for his work in educational and medical statistical analyses, including research and program development in areas such as multilevel modelling. In 1995 he was elected a Fellow of the British Academy and in 2007, he became a fellow of the Association for Educational Assessment, Europe.
Closing the achievement gap between Indigenous and non-Indigenous students is a critical policy focus in Australia. Simple analyses of the Australian National Assessment Program—Literacy and Numeracy (NAPLAN) data have shown that while some improvement in the performance of Indigenous students is occurring, the gap is still considerable. However, relatively little research has explored these data to answer complex questions. This presentation presents multilevel analyses of linked data on pupil achievement for a single cohort from Year 3 to Year 5 in Numeracy in Queensland, Australia, to explore Indigenous and non-Indigenous progress and variables affecting achievement. Results show the complexity of interactions in student achievement and the significance of prior achievement regardless of student background. Further, the importance of adjusting for measurement error in test scores to identify student patterns of performance is illustrated. The findings demonstrate that more appropriate statistical analyses can yield different implications for education policy. The importance of properly adjusting for missing data values will be illustrated.