Syndicate content

Friday links: computer-aided instruction, politics, reducing domestic violence, and more...

David McKenzie's picture

·         The IDB Development that Works blog covers a randomized trial of the one laptop per child program in Peru – no impact on learning, but some increase in cognitive skills.

·         The IPA blog also covers several studies of the impacts of computer-aided instruction in India.

·         A new paper questions the use of rainfall as an instrument for income in civil conflict studies.

·         A DFID evidence overview summarizes the results of impact evaluations of interventions designed to reduce domestic violence.

·         In Science this week, a paper summarizing the role of experiments in Politics – examples include list randomization to measure racial prejudice in the South, a study showing how political knowledge questions in surveys give much higher scores if people are paid per correct answer, voter turnout experiments, and more.

·         And in honor of daylight savings time this weekend, apparently we are all going to slack more on Monday according to this study.


Submitted by Paul on
OLPC. Calculations suggest the treatment group surpassed the control group by about 6-months of estimated learning. As implemented the program did not increase learning. ???? These seem like highly contradictory statements.

I'm not sure where you are getting this calculation from - the working paper does show impacts on cognitive skills, including a verbal fluency test, of about 5 months on average gain; but no increase in performance on language or math tests.