27 June 2014

How not to improve education in India

Some great analysis from MINT who highlight a new Government of India report, which ranks state education "outcomes".

What is odd is that the government rank has a negative correlation with the rankings of the Pratham report which directly measures learning outcomes.


So what goes into the government "outcomes" index?

- Number of teaching days
- Teacher working hours
- Enrolment rates
- Drop-out rates
- Primary-to-secondary transition rates

These are all basically inputs, with the exception of drop-outs and transition rates, which maybe say something about quality. But none of them are actually directly measuring learning at all. Yet more evidence for the Lant Pritchett case that focusing on inputs or "EMIS-visible" metrics won't get us quality learning outcomes, and measuring learning directly is critical to focusing policy attention on how to improve learning.

HT: Abhijeet Singh

2 comments:

Post a Comment