January 21, 2017 at 10:29PM

Last week, I wrote about the difficulties of educational research, and why we need more information on the circumstances in which research has been conducted.  After I had written that article, Daisy (Christodoulou) pointed out that E.D. Hirsch makes a related point in his article ‘Classroom results and cargo cults’.  Hirsch discusses a study which seemed to show that reducing class size improved educational outcomes for early years students and decreased the achievement gap between students from higher and lower income families.  The study prompted policymakers in California to spend $5 billion dollars on reducing class sizes in their state. Despite the injection of funds and the apparent evidence that reducing class size worked, the results in California were disappointing.  Hirsch argues that the reason that they were disappointing was because the researchers who undertook the original study did not provide a ‘theoretical interpretation of [their] own findings’.  They did not, for example, include any systematic analysis of why smaller class sizes seemed to benefit younger students more than older ones.  The policymakers in California who read the study therefore concluded that reducing class sizes was sufficient to achieve the gains they sought.  They did not ask what else they might need to do alongside reducing class size because no one had suggested to them that this was an issue that they needed to consider. This leads Hirsch to conclude that studies which report improvements in educational attainment should be accompanied by an academic exploration of the precise causes of these results.  I can recommend reading Hirsch’s article in full.  It can be found here: http://ift.tt/2jPcxIu.  Anyone interested in how we use evidence in teaching should also read Daisy’s article on why the evidence from randomised control trials, although useful, is less useful than scientific evidence on how the brain works:  http://ift.tt/2jF4yuZ.





from razorbladeinthecandyfloss http://ift.tt/2jP2H9M