Harvey Goldstein, who has died of Covid-19 at the age of 80, has left a formidable legacy from his work, both as a statistician and as a campaigner for more careful scrutiny of assessment data in education – whose misuse he consistently queried.
Harvey’s career included posts at the UCL Institute of Child Health (ICH); as Professor of Statistical Methods, Institute of Education (IOE), University of London, 1977 – 2004; and as Professor of Social Statistics at the School of Education, University of Bristol, 2005-2020, where he remained working right up until his death.
He represented a rare combination of statistical insight, rigour and inventiveness, coupled with a fierce desire to call out the abuse of data in public debate and broaden conceptions of what evidence-informed policy should really look like. In all these ways he was a committed activist, as well as an academic, applying his insights to practical problems in education that would benefit from such close scrutiny. His 2019 Otto Wolff Lecture, “Living by the Evidence”, given at the ICH to celebrate his 80thbirthday, shows many of these qualities.
It is indeed an impassioned plea for public policy decisions to be influenced not just by the research evidence but also by “priorities, feasibility, acceptability, ethics” which he considered to be matters of judgement, and inevitably, therefore, political.
Doing research in this way meant being engaged. Indeed, more than most academics he had a sustained track record of making a difference by the problems he chose to work on and the way in which he interacted with policymakers to try to bring about change. The thorny issue of school rankings and what he saw as the misuse of performance data as crude accountability measures played a central part in his work in education.
While at the IOE, Harvey established the Centre for Multilevel Modelling (CMM), which moved with him to the University of Bristol in 2005. CMM has been pivotal in advancing the application of statistical method to complex problems in the social sciences, and education in particular, not least through its development of multilevel software, dedicated to working on hierarchically structured datasets.
Through the Centre’s development, his own publications, as a member of the Royal Statistical Society (RSS), the International Statistical Institute and Fellow of the British Academy, and as joint editor of the prestigious RSS journal Series A, Statistics in Society, Harvey brought on a whole generation of statisticians, many of whom now hold prominent positions in their own right.
Throughout he was determined to devise practical solutions to some of the dilemmas educationalists faced and argue for them. In this spirit, the Otto Wolff lecture sketches out an alternative approach to the overuse of testing in schools right now. Amongst those who will miss him most are the British Educational Research Association expert panel, convened to work with him on this idea and develop it into a practical proposition for educationalists, parents and politicians to consider. We will carry on. As Harvey himself said in his Otto Wolff Lecture:
“The important thing for researchers is to not give up. The research and the publicising of the implications of that research, along with public critiques of evidence abuse or suppression, need to continue. All of this is difficult, but I think there is an ethical imperative to try to do it. And I hope to be involved in doing just that.”