Sunday, May 15, 2016

A summary of latest evidence on education research

Simon Burgess has an excellent compilation of all the latest research on various aspects of human capital formation and education. No surprises anywhere. Nevertheless, a few of the important findings copied below. 

1. Woessmann highlights a set of five practices of effective schools, from close analysis of high-performing charter schools. Thet are “increased [teaching] time, better human capital [i.e. more effective teachers and administrators], more student-level differentiation [i.e. supplemental tutoring], frequent use of data to alter the scope and sequence of classroom instruction, and a culture of high expectations”. 

2.  What creates an effective school system? Accountability, even relatively low stakes one, which requires some common and consistent form of assessment (generally centralized exit exams), when coupled with operational autonomy for schools, have been found to improve pupil attainment. But it is important to ensure that the assessment is well designed and captures the skills that society wants pupils to have, and does not end becoming the standard case of "teaching to the test". 

3. On the same line of analysis, we need to have more research on what effective teachers do that makes them effective. In particular, how to structure, implement and incentivise take-up of effective initial teacher training and continuing professional development, as well as a pre-hire screen that is usefully selective on effectiveness. 

4. Publication of rankings showing schools are better performing can in principle increase socio-economic sorting of pupils, though the evidence on this is mixed. It is also not clear as to whether it does so depends on the admissions process to schools and whether this is manipulable by parents.

5.  While there are a few robust studies showing positive effects of higher resources for schools and lower class sizes, there are more studies showing little or no effect of class-size on outcomes. So inputs, especially in countries like India, should be less of a focus area. 

6. There is little strong evidence that adding more IT hardware to classrooms raises attainment. There are stories of what brilliant teachers can do with IT, but brilliant teachers raise skills without IT as well. This does not seem a cost-effective education strategy.

7. Non-cognitive skills is an area of which very little is known -  a better understanding of, and measurement system for, non-cognitive skills; a better understanding of how the cognitive and non-cognitive skills discussed above map into workplace skills (such as knowledge skills, problem-solving skills, critical thinking skills, and social skills) and how they are valued 

8. Finally on vouchers, the jury is still out. There are two main research and policy questions: what is the impact of the voucher on the individual who receives it? And what is the impact on the system as a whole, on those ‘left behind’ in the low-performing schools? There appear to be no definitive answers to the two core empirical questions yet. In a substantial recent review, Epple et al argue that the bulk of the findings suggest no significant effect, yet “multiple positive findings support continued exploration”. The evidence on the impact on the voucher-using student from the US is mixed, though there are positive effects on non-academic outcomes and the reason that parents enter school lotteries rather than for attainment improvements.

1 comment:

Karthik said...

This is great but I believe such research should be presented in a different way. Presenting it in form of laundry lists which most meta-analyses papers or papers reviewing literature tend to do isn't useful and some times has negative consequences.

My post on this here on arguments against laundry list form of communicating and proposing a new way to communicate research.