- Chris Blattman provides an incentive to delay giving up on that great research idea you’ve been peddling for years in this story from the EconTalk podcast: For years, he pitched random African factory owners the idea of an RCT of factory employment. “They’d usually look at me kind of funny. They wouldn’t leap at the possibility. I was just this person they met on a plane.” One day it worked, and six weeks later he was randomizing applicants.
- Andrew Gelman on the “what does not kill my statistical significance makes it stronger” fallacy, and why even Heckman has fallen victim to it. “conditional on statistical significance at some specified level, the noisier the estimate, the higher the Type M and Type S errors. Type M (magnitude) error says that a statistically significant estimate will overestimate the magnitude of the underlying effect, and Type S error says that a statistically significant estimate can have a high probability of getting the sign wrong”
- Let the data talk, but what language should it speak? 25 ways of visualizing the same dataset on life expectancies around the world.
- In the latest JEP, how to write an effective referee report: With three specific recommendations: I) Make clear the contribution, and give appropriate value to innovative work: “The importance of a contribution can be undervalued in some cases by referees and editors. After all, papers that are more ambitious are often more likely to have loose ends, which gives referees and editors a reason to avoid taking a chance on them” II) divide comments into two clearly demarcated sections: 1) problems that make the paper unpublishable, which (if revision is invited) must be addressed before the paper is publishable; and 2) problems that are not essential for the publishability of the paper, which should labeled as “suggestions.”; and III) In making requests of authors, weigh the costs of the request. It is not enough that a particular request will improve the paper. The benefits must exceed the costs, so that the improvement has positive net present value. Since the author bears the costs, it is easy for a referee to make absurd demands thoughtlessly. Don’t.” – and finally, after receiving multiple 5+ page referee reports recently, I agree with “Unless a referee needs to make extremely technical points, 2–3 pages should be sufficient.”
Cash transfers are great – lots of people are telling you that on a continuous basis. However, it is an open question as to whether such programs can improve the wellbeing of their beneficiaries well after the cessation of support. As cash transfer programs continue to grow as major vehicles for social protection, it is increasingly important to understand if these programs break the cycle of intergenerational poverty, or whether the benefits simply evaporate when the money runs out…
- In Science last week, Rema Hanna summarizes several studies that look at the ways technology such as biometric smartcards is helping to reduce corruption in transfer programs.
- Marc Bellemare has a second post on dealing with imperfect instruments.
- The Statablog on how to use the command putexcel to export your Stata output into nice fancy tables.
Impact evaluations of interventions aiming at reducing intra-partner and sexual and gender-based violence (IPV-SGBV) have mostly failed at detecting statistically significant impacts.