Two academic heroes under 30

I have two academic heroes this week.

The first is U Mass graduate student Thomas Herndon, who replicated Rogoff and Reinhart's austerity paper and discovered a mistake.  Many social scientists encourage replication, but people arguably don't do it enough.

The second is a public high school sophomore Ria Chhabra who began performing experiments on fruit flies during middle school, and discovered that organic food results in better outcomes for fruit flies. 

Their examples can inspire and encourage all of us.  Their achievements are an important reminder of how capable students can be, given opportunity and encouragement, and how it doesn't take elite schools to inspire achievement.  And how important it is to question assumptions, design good work, and work hard to execute it. 

Emerging adulthood and medical subject headings

Emerging adulthood has been recognized as a concept in sociology for over a decade, and is generally recognized as extending to age 30. Adolescent health research often extends up to age 25 because brain development continues until about that age. Last year I asked Alan Guttmacher, the head of the National Institute for Child Health and Development (NICHD), about the definition of adolescents for the purposes of NICHD funding, and he said that it's reasonable for research to go past age 18.

In view of that, I'm submitting a new manuscript today, and as I was choosing keywords, I was surprised to notice how they classify ages.

Adults are ages 19-44. Young adults, a term added in 2009, are ages 19-24.

My study is about the transition to adulthood --- how people get college degrees in early 20s and develop professionally in their late 20s and early 30s --- and it doesn't seem to fit into either of these categories. The study examines how the subjects are in the process of becoming adults, socially and economically, but according to the medical subject headers, they are already. It's striking to be reminded of how our systems of thought have changed in such a short time, and an indexing system naturally didn't catch up.

Publication bias and the result that got away

An economist writes about his would-have-been dissertation: he gathered the data, did a regression, and found no relationship. So his dissertation was on another topic. 8 years later, he publishes the null relationship. It's refreshing to see someone talking about how publication bias affected their own research, as opposed to the studies published in epidemiology journals that suggest publication bias in macro.

Exam strategies and meta-education

The humor site Cracked has a good exam study strategies article. When someone posted this piece on facebook, several faculty members mentioned that they often accumulate more materials than they read.

How many students learn how to learn? Would we have better exam outcomes if we taught students explicitly how to study?

I recall that research shows that such study strategy courses pay off in students doing well in their other courses, but still few students have the opportunity to take such meta-education courses: courses in how to study, learn, and benefit the most from coursework. Instead, students are left on their own to figure out by trial and error and possibly wasting the courses where they do less well because they aren't sure what to do. I know students find these types of courses to be tedious, or they feel like they ought to already know this material, so they act blase about learning the material. I remember feeling that way about Freshman Advisory class in high school, where they taught us how to study before our first exams as we rolled our eyes. The usual tips like don't study in front of the TV or in bed, but the tips stayed with me and every time that I tried to study in bed, I would remind myself that probably it would be a better idea to move to my desk. Although I wouldn't always.

Cognitive dissonance in faculty recruiting

The University of South Carolina has what I understand to be a very high quality School of Public Health. They are looking for a tenure-track faculty member with a description that matches my own research fairly well, studying "sexualities, sexual behavior, sexually transmitted infections, HIV/AIDS, teenage pregnancy, LGBT disparities or other sexuality and health issues." Particularly, they want someone who advocates and engages with community, state, national, and global organizations.

I checked their sex ed laws, and according to this profile, they are one of 7 states in the US to prohibit positive depictions of homosexuality in public schools, not a favorable start for any researcher who studies LGBT issues to advocate or engage with the community and local institutions. They also prohibit contraception information other than in the context of future family planning, which is better than nothing. SC is also one of the minority of states to apply for the federal abstinence-only sex education funding, meaning that the programs funded by that funding cannot even mention contraception in any positive way at all.

On the bright side, they also applied for the federal evidence-based sex education and pregnancy prevention funds, which they could certainly use since SC has the third highest gonorrhea rate in the country, and 8% of women ages 15-19 get pregnant each year, and 5% give birth. Looking at the gap between pregnancy and birth rates, it's particularly remarkable because 72% of SC women live in a county without an abortion provider (93% of counties lack an abortion provider), Medicaid does not fund abortions, they mandate a 24 hour waiting period and all procedures 12 weeks and after are a felony with mandatory minimum of 5 years in jail and/or $5000 fine.

Thanks to NARAL and Kaiser for the facts. The SC ACLU also has a long access guide listing all the sexuality, family planning, and related resources (e.g., PFLAG) in the state and neighboring states. So the state fortunately has some resources, even if they get an F by NARAL.

Attractiveness ratings

A Psychology Today article about attractiveness ratings in the Add Health data has yielded some interesting responses: that it's wrong offensive inanity and apparently not unprecedented awfulness that can't be replicated in reanalysis and could cause the author to join the Bell Curve author at a conservative think-tank and become an anti-PC crusader.

One aspect that I've not seen discussed is whether the author has anything in common with his research targets given that in other contexts, it's been widely noted that both African-American women (the target of the original article) and Asian men (like the author) are marginalized in the dating marketplace: e.g., on this African-American dating website, OK Cupid's data analysis, and a roundtable where I presented at American Sociological Association included a paper about how gender stereotypes hurt Asian-American men because Asians in general are considered "feminine" and African-American women because African-Americans in general are considered "masculine": hyperfeminine women and hypermasculine men are considered desirable, but feminine men and masculine women are not. Separate research documents how Asian men may feel emasculated, and allegedly business research documents issues with incorrect condom sizes in Asia.

Marginalization is a well-known explanation for bigotry, although of course not an excuse.

Would-be position

My faculty homepage at Stony Brook Medical School, which cancelled my position due to NY State funding cuts. Sigh.

Quantitative work is not mysterious

Some are impressed by quantitative skills. Those of us who were math and physics majors (or in my case, both) are more realistic. Monday's xkcd gives a secret trick in the mouse-over box that pops up that all math and physics majors learn. For those who haven't figured it out, I'm glad that there's a comic strip to reveal the secret.

Also, a few weeks ago, there was a good xkcd about the attitude of some science types towards about interdisciplinary work.

White males are perceived as more competent

Teaching evaluations for statistics professors are almost always lower for women than for men. There's a similar finding by researchers at U Wisconsin Milwaukee that customers in general are more satisfied with the performance of white men than minorities or women.

The story shows the white male researcher's picture, giving them the opportunity to do a meta-study to find out whether including the picture makes the reader more convinced of the correctness of the findings.

Funding transformative research

NY Times article on conservatism in science funding. To add to the chorus: it's not curing cancer or anything, but my virginity pledge research would have probably not gotten funded by NIH. I didn't expect to find that virginity pledgers would lie about having had sex, so I wouldn't have known to look for that and I wouldn't have been able to write a proper research proposal to do that.