Dev.SN ♥ developers
Sections
Dev.SN
from the this-will-not-be-controversial-oh-no-sir dept.
GungnirSniper writes:
"Catherine Rampell at The Washington Post has 'A message to the nation's women: Stop trying to be straight-A students.'
In her analysis of others' findings, she writes of a discouragement gradient that pushes women out of harder college degrees, including economics and other STEM degrees. Men do not seem to have a similar discouragement gradient, so they stay in harder degree programs and ultimately earn more. Data suggests that women might also value high grades more than men do and sort themselves into fields where grading curves are more lenient.
'Maybe women just don't want to get things wrong,' Goldin hypothesized. 'They don't want to walk around being a B-minus student in something. They want to find something they can be an A student in. They want something where the professor will pat them on the back and say "You're doing so well!"'
'Guys,' she added, 'don't seem to give two damns.'
Why are women in college moving away from harder degrees?"
(Score: 5, Insightful) by Sir Garlon on Wednesday March 12 2014, @04:59PM
This is an interesting correlation that might actually bear some further examination. Briefly, the observation is
Unfortunately, the research was done by an economist and seems to stop at pointing out a correlation as if that told us anything about causes. The logical next step, in my opinion, is to do some experiments to determine what, if any, causal relationship is represented by the correlation.
From reading the Washington Post article, the reporter seems to be trying to imply that women are less tough-minded then men and give up quickly when they face the hard uphill battle that is Econ 101. That's just one possible explanation, and probably a stereotype. Another possible explanation is that women make up their minds in Econ 101 whether economics is really for them, and those who decide "no" apply themselves less after they decide they prefer another major, whereas men are more likely to have made up their minds before they enroll in the class. Or maybe men are just less quick or less likely to change majors. Or maybe economics professors are more likely to be misogynist than literature professors. I am not saying I believe any of these explanations, but I could probably come up with 10 of them in an hour.
The problem I see in these studies of women in STEM is that generally the researchers notice some correlation or pattern like this, and somehow assert they've learned something. Without a hypothesis as to the underlying cause, and research to test that hypothesis, *we don't even know if this pattern is a problem.* It could be women turn away from economics because of negative factors that push them out (which would be bad, and should be addressed), or it could be because of positive factors that attract them to other subjects (which would be no problem).
[Sir Garlon] is the marvellest knight who is now living, for he destroyeth many good knights, for he goeth invisible.
(Score: 3, Insightful) by Daniel Dvorkin on Wednesday March 12 2014, @05:45PM
Awfully hard to set up a controlled experiment for this kind of thing. Probably the best you could do would be a prospective study: start with a bunch of freshmen of both sexes and track their grades, their attitudes (as measured by some standardized questionnaire), and their majors throughout their college careers. You could, at the very least, establish temporal relationships this way, which would help with untangling the no doubt complicated relationships of cause and effect.
Pipedot [pipedot.org]:Soylent [dev.soylentnews.org]::BSD:Linux
(Score: 1) by dcollins on Thursday March 13 2014, @12:30AM
"The logical next step, in my opinion, is to do some experiments to determine what, if any, causal relationship is represented by the correlation."
Agree with Dvorkin... I suspect you don't entirely know what you're saying. What kind of experimental design would you recommend? Randomly inflating some people's grades beyond what they earned to see if they stick around longer? That's infeasible.
Sometimes correlations are useful information -- and sometimes that's the best you can do under ethical restrictions.
(Score: 2) by Sir Garlon on Thursday March 13 2014, @09:58AM
As Dvorkin said, it's hard (probably futile) to attempt to design an experiment to determine the cause of human behavior. Heck, it is probably hard to do that even for rats in a laboratory. So "experiment" is perhaps not the right word. What I would recommend would be to formulate several causal hypotheses and then investigate each and see which, if any, have statistically significant observational support. You could survey the male and female students when they exit Econ 101 and ask them how hard they thought it was on a rough scale from "very easy" to "very hard." (hypothesis: women are more likely than men to think the course was hard.) You could also ask other questions like how well prepared they felt they were when they started the course (hypothesis: people who quit economics did not meet the assumed prerequisites), whether they thought the professor was helpful and approachable (hypothesis: professor is perceived differently by women than by men), how they're doing in their other classes (hypothesis: econ majors who do worse in economics than in their other classes are more likely to change majors), and things like that. Then analyze the responses in comparison with the students' grades, and see what light this sheds on the gender differences. Repeat for different economics professors, over several years, at different universities. Let the breadth and duration of the study be dictated by how much data you need to make a defensible conclusion, instead of what data you can get cheaply in time for the next conference.
I recognize that a survey is not an ideal instrument for investigating people's motives, and that there is US law that may restrict the researcher getting access to the students' grades. But this are just the ideas I've come up with in casual conversation, and perhaps a psychologist or social scientist could come up with some better instruments with which to gather the relevant observations.
[Sir Garlon] is the marvellest knight who is now living, for he destroyeth many good knights, for he goeth invisible.