[saint_fiasco]

There were studies made in India that showed otherwise. Without segregation the teachers were teaching to the level of the smartest students in the class, so the slower students were being left behind and became unmotivated. With segregation, the teachers were still teaching to the level of the smarter students in the class, but that level was much closer to the level of the slower students so they could catch up.

Keep in mind the actual reality of the situation might be different in each country, so it's not like one set of studies was necessarily wrong. Maybe teachers in the US teach to the level of the average student, so the results would be different there.

[bayesian_horse]

The assumption there is that the ordering of the students by performance is stationary, particularly in the first few years. Regardless of personal situation.

If a child catches a flu at the wrong moment, it will get stuck way below its level for all its career.

Khanacademy data has shown that students in a math class will progress at different speeds at different times and that their "ranking" can invert multiple times over a school year or beyond. It's not exactly the same, but their system makes sure students don't stay stuck.

[adrianN]

That's really interesting. Do you perchance have something that might help me find those studies?

[saint_fiasco]

I found them on a book called "Poor Economics". The authors also teach an online class on development economics where among other things they talk about education in poor countries.

https://www.edx.org/node/92491