An Experiment: Paying students for good test scores

[tectonic_drift]

cross posted to policyresearchA school district in Ohio is examining whether offering students monetary awards for performing well on tests improves test scores. The district is using a randomized experiment to investigate the question. The results are due out in June, but an overview of the study and the efforts it took to get schools and

Comments

[nyuanshin]

Such a finding wouldn't really support your position (but it wouldn't support its converse either), since it's exactly what straightforward economic reasoning would predict. Both you and the people who are in favour of this program should expect there to be a statistically significant performance difference. If there wasn't, it'd be Gary Becker who'd have something to worry about.

The econ nerd in me likes this sort of thing, but taking a more holistic view this is clearly a kludge of a solution to the problem of kids not wanting to do what their teachers and parents want them to do. I understand where you're coming from, but your position is flawed for the same reason the opposing position is: we need to take into account the underlying variation in aptitudes and motivations of children. Kids are not widgets. Some will benefit from the kind of incentive structure you favor, others will benefit more from a monetary one. Focus trying to find One Right Way in education obscures the greater need for pluralism in education.

[tectonic_drift]

You are right to point out that observing a decrease in performance (after rewards are removed) would support the program hypothesis that incentives matter. And would only re-enforce my concern about motivations for learning….but my interest in the topic was more about the experimental methodology and thinking about incentives.

I definitely agree with you that there is no one-size-fits-all solution to improving education. People will respond differently to the same incentives. Ultimately, the responsibility for education is shared between parents, children, and the schools system. And I sometimes get frustrated with the idea that more money is needed to fix the problem.

[nyuanshin]

Agreed. "More money" is what's been tried for the past 30 years running, clearly without significant effect. We need more of this sort of experimental approach to education

[tectonic_drift]

I essentially agree with what you say. I like how you phrased it: kids are not widgets.

My main hesitation of the voucher model for public school is that in other instances the market has abandoned some neighborhoods. For instance, many high poverty-high crime neighborhoods in the U.S. are without grocery stores or banks within in a reasonable distance (i.e. the average distance in the suburbs). Private grocery stores and lenders do not see it in their interest to set up shops in these areas (understandably so). I am not opposed to vouchers, but just hesitant and cautious.