Guest post by Jeff Mosenkis of Innovations for Poverty Action.
- The new Freakonomics podcast, “What are gender barriers made of?” is a nice look a little deeper than surface statistics at the subtle behavioral factors responsible for gender gaps in the labor market. They look into research on non-conscious differences in the ways people listen to men and women, and point to a few things we can change right now that we take for granted – interviews and performance review self-ratings that have been shown to have little to do with performance but tend to be biased in favor of men.
- Economics is one of the few fields to list authors alphabetically, rather than as a signal of who contributed most to the paper. A few months ago Heather Sarsons found that women in economics face a “co-authorship penalty” in their careers (presumably, alphabetical listing leads to ambiguity and allows bias to creep in). Now a new paper reviews evidence for “alphabetical discrimination” in economics. Researchers with last names later in the alphabet (more likely to become “et al”) get tenure at top departments less, win fewer top awards, and their papers are downloaded less, but also respond strategically.
- To summarize: economists’ revealed preferences suggest they prefer less information in their own labor markets.
- Ashraf, Bau, Nunn, & Voena have a new paper suggesting the bride price tradition of a groom’s family pays the bride’s family for her may have an upside. In Indonesia and Zambia they find that more educated brides have higher bride prices, and when new schools are built, ethnic groups with bride price traditions increase girls’ enrollment. (h/t @DinaPomerantz)
- Researchers have found a statistical glitch in the popular fMRI analysis software packages, allowing them to find false positives up to 70% of the time, which calls thousands of study results into question (usually the type behind the “when you do X this part of your brain lights up” headline stories).
- Resource: a roadmap for conducting systematic reviews, dealing with confounding variables and other tip sheets in English and French (h/t David Evans)
- An interesting long read from the Science of Us about how it took years and several incorrect published papers for researchers to admit they had their correlation sign backward.