Thursday, July 12, 2012

TL;DR - Relative Benefits From Religion Is being religious a good thing?  Previous research has found that people who are more religious tend to have better social relations, psychological adjustment, and even better health.  Gebauer and colleagues, however, found that the strength of these correlations can vary depending on the country a person resides in.

For the study, they used people's responses to questions for a popular European dating website, eDarling.  (Note that while this is self-report data--and self-report data on a dating site, at that--it's unlikely that there are going to be any systematic biases in reporting by country.)  The average correlation between religiosity and social self-esteem or psychological adjustment is on the order of .1 (not huge, but as a general rule there are no huge effects in social psychology).  This varied quite widely across the countries looked at, though, with more atheistic countries such as Sweden and the Netherlands having essentially no correlation, while the much more religious Turkey had a correlation of .2. 

So maybe it's less important how religious you are, as much as how culturally normative you are.

Jochen E Gebauer, Constantine Sedikides, & Wiebke Neberich (2012). Religiosity, Social Self-Esteem, and Psychological Adjustment: On the Cross-Cultural Specificity of the Psychological Benefits of Religiosity Psychological Science, 23 (2), 158-160 DOI: 10.1177/0956797611427045

TL;DR - This may be my last post...

Psychological Science, February 2012 -- "Saving the Last for Best: A Positivity Bias for End Experiences," Ed O'Brien & Phoebe C. Ellsworth.

...Okay, probably not, but maybe now you'll like it better.  We've known for a while now that how something ends can alter someone's evaluation of it (c.f. Redelmeier & Kahneman, 1996), but O'Brien and Ellsworth found that simply highlighting the fact that something is the last makes people appreciate it more.

Wednesday, July 11, 2012

TL;DR -- She's Just Not That Into You

Psychological Science, February 2012 -- "The Misperception of Sexual Interest," Carin Perilloux, Judith A. Easton, David M. Buss.

As every human knows, it can be hard to know when someone likes you.  We end up devoting a lot of cognitive effort (and song lyrics) to figuring out who would like to have sex with us, and dealing with the consequences of these beliefs.

In this study, Perilloux and colleagues set up a speed-dating situation, while also having each participant provide various ratings (including attractiveness and sexual interest) of each person they met, as well as self-ratings.  This way, they were able to contrast each partner's perception of the other and see how well they matched, or mis-matched. 

They found that men tend to overestimate the sexual interest of women, but that women tend to underestimate the sexual interest of men.  Also, the more attractive a man is, the worse his estimates of interest get, this despite the fact that women tend to also have more genuine interest in more attractive men.  When a woman is rated as more attractive, she's more likely to be misperceived as displaying interest in a man.

So far, so good.  This is a pretty interesting finding, and has implications for understanding complex behaviors such as rape.  However, the researchers try to explain the findings using an evolutionary framework, which I find unnecessary. 

Objection number one: they looked exclusively at an undergraduate sample from the University of Texas (mean age: 18).  You could argue that, as people in their reproductive primes, this sample should be most representative of our evolutionary heritage.  But still, American undergrads are poor representations of the majority of humans on the planet, living or dead.  I always like to see some extensions to different populations before I'm comfortable with an evolutionary hypothesis.

Objection number two: the evolutionary theory isn't needed.  For instance, they try to explain why more attractive women would be more likely to be misperceived as interested by attributing this to an evolutionary bias based on the correlation of fertility and attractiveness.  But this could also be explained by, essentially, wishful thinking.  Our desire for something to be so can influence our estimation of its likelihood.  A cognitive explanation can substitute for an evolutionary one.  And likewise with the difference in over-/under-estimation between men and women: this could be a difference in set points based on the level of sexual interest in the individual him/herself.  Or the finding that more attractive men overestimate more (over-overestimate?): well, as it's pointed out, they also have more success than less attractive men, so could this just be driven by learning?  More frequent reward leads to greater perception of interest, more frequent punishment leads to less?

The authors need to do a better job setting up the kind of background necessary for an evolutionary explanation to be sensible.  Is accuracy of perception of sexual interest linked with differential fitness?  It's not really clear; it's briefly alluded to and glossed with citations -- all of which are connected to one of the authors (Buss).  But this is a prerequisite for making a claim that a trait has evolved!

Overall, there's nothing in this study that couldn't be explained through more fundamental cognitive or learning mechanisms.  Does that mean an evolutionary explanation is wrong?  No.  It just needs more convincing evidence.

Wednesday, July 4, 2012

TL;DR - Teacher Experience Matters

Science, March 2, 2012 -- The Effects of Experience and Attrition for Novice High-School Science and Mathematics Teachers, Gary Henry, C. Kevin Fortner, & Kevin Bastian

There's a lot of things wrong with the public education system in the US, but one of them is its ability to retain high-quality teachers.  Due largely to anemic rates of compensation, a student in public school is likely to encounter either a bright, motivated young teacher -- who disappears in a few years -- or an older teacher who may or may not be that effective.

What does this system mean for students, though?  We already know that teacher quality can play a large role in determining student success, but how much does teaching experience contribute?  If we just look at the number of years that a teacher has been around, we definitely see a positive correlation with student success.  However, is this due to the experience gained by those more senior teachers, or is it just an artifact of less competent teachers dropping out of the profession earlier?

Henry, Fortner, and Bastian attempted to address this question in a large-scale retrospective study looking at teachers in North Carolina.  Their sample included end-of-year test scores for hundreds of thousands of students, linked with thousands of teachers across several math and science (STEM) courses, as well as several non-STEM courses.  In the analysis, they were able to separate and compare the teachers who would and would not continue teaching after five years.

One thing they found was that teachers who remained past five years were more effective than teachers who dropped out earlier, supporting the hypothesis that lower-quality teachers leave the profession earlier.  But they also found a clear upward trend over the years for the other teachers; over the course of five years, the test scores for the teachers' classes increased reliably.  Experience does matter.

But all experience is not equal.  While there was a general upward trend for all teachers, physics and chemistry teachers demonstrated the greatest increase.  How about those non-STEM teachers?  They actually showed much smaller gains in efficacy over the years, essentially leveling out by year 3.  It should be pointed out that this doesn't mean the non-STEM teachers were any worse (or better) than the STEM teachers.  Comparing physics scores to English scores is one of those apples/oranges situations.  So it's not clear why there was this difference across disciplines.  Is there something about STEM courses that makes hands-on experience more valuable than for non-STEM courses?  Or is there something about the teacher training that the teachers receive?  Unfortunately, we can't really tell from this study.

Regardless, it's clear that the sort of rapid turnover that we're seeing in public schools is doing a big disservice to our students.