From: The Australian 28/6/2017
Author: Darragh O'Keeffe
The emphasis on socioeconomic status in education research and policy is unwarranted as it does not have a strong relationship with academic outcomes, new research claims.
The analysis by the Australian Catholic University's Gary Marks is the latest contribution in a longrunning debate over the influence of a child's background on their education outcomes.
In a paper published in Australian Educational Researcher, Dr Marks argues that SES has a "particularly weak" relationship with education when stronger influences such as earlier achievement and cognitive ability are taken into account.
Dr Marks analysed data from two waves of the Longitudinal Study of Australian Children, which included years 7 and 9 National Assessment Program Literacy and Numeracy scores and measures of family income, parental education and occupational status.
While there is a "pervasive view in academic publications, commissioned reports and commentary that SES is the most important driver" of outcomes, Dr Marks said individual measures such as family income, father's occupation status and parents' education were at best only moderately associated with achievement. Composite SES measures, where two or more indicators are combined, had only slightly stronger correlations with achievement, he found.
While previous research has concluded that school-level SES matters greatly for achievement, Dr Marks argued there are "serious statistical problems" with this measure. "When taking into account prior achievement or unobserved differences between students, school-SES effects are negligible," he wrote.
Area-based measures of SES, which the Gonski model recommends be used to allocate funds until a better measure is developed, show weaker associations with student achievement than student-level indicators, he said.
"Given the weak associations between area-based measures of SES and student achievement, funding models that rely on these measures are unlikely to be successful," he argued.
Dr Marks' analysis found cognitive ability and earlier achievement had a much stronger relationship with student performance than SES. He says researchers and policymakers should be mindful of the weak effect of SES, the stronger effect of other factors such as prior achievement, the smaller than assumed between-school differences and the "sizeable genetic component" to outcomes.
Ian Li, an economist at the University of Western Australia, agreed that SES, and in particular the measures used widely in research and policymaking, "are not optimal". He also agreed that earlier academic achievement was a strong predictor of subsequent outcomes but was not convinced that SES did not influence earlier academic performance.
"This reinforces my perspective that early childhood intervention is important and could potentially pay large dividends in terms of improved outcomes across the life cycle," Dr Li said.
Equity measures needed to be targeted early in life for disadvantaged people, he said. "It has been suggested that disadvantage starts early in the life course and its influence is so pervasive that real action needs to begin in early childhood and perhaps even before birth, by improving outcomes for parents and family."
Jenny Chesters, a research fellow at the Youth Research Centre at University of Melbourne who has analysed NAPLAN data in the ACT, said her work "clearly shows SES has an effect in each year". "Even after controlling for NAPLAN score in Year 3, SES has an effect on NAPLAN scores in years 5, 7 and 9," Dr Chesters said. "Family circumstances have an effect on the level of school readiness, which in turn affects standardised test scores in Year 3.
"Furthermore, attending a school with a high proportion of low SES students has a negative effect on NAPLAN scores, even after controlling for family SES."
A major data analysis in 2015 by the Mitchell Institute found "socioeconomic disadvantage has a greater impact on educational opportunity than any other factor" considered in that study. The research found a "clear and persistent relationship" between SES and education outcomes across four milestones: early years, middle years, senior school and postschool.