warw@uunet.UU.NET (Anthony R. Wuersch) (03/17/89)
Re the Time and New York Times articles on a gender gap in computing,
does anyone have this article or book that interested the journalists?
I'm wondering about the usual sociological questions that get asked of
a study, i.e.,
1. what is the research design,
2. what biases does it admit to having,
3. what alternative hypotheses does it respond to,
4. what were its dependent and independent variables,
5. what kind of data did it use --- survey questions, census data, ....
6. if there was a sample, how was it selected,
7. if there was a survey, could respondent answers be cross-checked with
data that didn't come from the respondent?
8. what were the results, and how much of the gap was explained?
etc..
A gross example of 2. is selection bias, i.e., using only computing
people and including no non-computing people in the study.
An example of 7. cross-checking is math and science club membership
records to check for scientific bent. A nice study might see if people
who show an interest in science while in high school (via membership)
end up in science, and at what rates when compared with the general
population.
Point 8. is also important. Only the most successful of such studies
explain more than 40% or so of variance. That means for most cases,
the researchers don't know.
Many people who read journalists expect that a social scientific study
explains what most people do. Usually, no. It has the most meaning
for policy makers who want to know where they should invest money.
It can also help explode popular myths.
I don't recall any decent work on the relationship between success in
US adolescent peer groups and career choices. American high school
experience, at least as far as role models and media treat it, puts
career choice far below college choice in importance. That was my
feeling in high school.
In northern continental Europe, many students are tracked in high
school on scientific, humanistic, or vocational paths on the basis
of testing and inclination. Their future choice of career is more
restricted, but the scientists don't seem to go through any social
crises of rejection or isolation which drive them on to science.
Many more engineers and scientists come out at the end. The school
system can also help decide how many women will enter the scientific
track. I think in more liberal countries, it does help.
Toni Wuersch warw@cgch.UUCP
Ciba-Geigy AG ..!mcvax!cernvax!cgch!warw
Basel, Switzerland