What's going on here? PART TWO

Mike Salovesh (t20mxs1@CORN.CSO.NIU.EDU)
Fri, 26 Apr 1996 07:13:41 -0500

My previous message took a lot of space to set the stage. Now, at last,
here's the problem, or anomaly, or whatever it is that turned up on the
last test I gave to 117 students in my introductory anthro course:

Something seems to have led many of my students astray when they tried to
handle any question that dealt with matters of sex and gender. Related
questions that were not clustered in any one place within the exam either
had very low or actually had negative correlation coefficients: the LOWER
a student's score on the test as a whole, the more likely it was that
student would get one of these questions RIGHT. Again, I try to
design questions so that on average about half the students in the class
will get any particular question right. Even so, I don't expect that
any one WRONG answer to a particular question will attract a
heavy majority of the class, and that is what tended to happen on these
questions. Let me give some condensed examples, all of which turned out
to have either around zero or even negative discrimination indexes:

Among all the primates, humans exhibit A) the geatest sexual dimorphism;
B) the least sexual dimorphism; C) the least amount of plasticity in sex
roles; D) the greatest amount of plasticity in sex roles.

Results: More people marked A than D. But these folks saw several videos
of living primates -- and at very least should have remembered something
about the huge canine teeth in male baboons. Yes, I talked about this in
class, too; and it's in the text, besides.


The facts involved in most common cultural definitions of what it means
to be a male or a female in the U.S. today are A) independent of the
facts of biological differences; B) not the same as the facts of
biological differences; C) determined by the facts of biological
differences; D) not related to the facts of biological sex, as such.

57% of the class marked response C! (A, 15%; B, 21%; D, 8%) Now what
the hell? I can hear what I said about forty-'leven different ways, and
what I said was that biology is NOT destiny, but it does have something
to do with it because there is a widespread cultural belief that it makes
a difference.


What kind of relative is Ego's FaFaBr? A) A patriarchal relative;
B) An affinal relative; C) A lineal relative; D) A collateral relative.

46% marked A, 10% marked B, 30% marked C, and 14% marked D. And I
thought I had talked myself to death trying to point out that patriarchy
and matriarchy are terms that refer to politics, not to the
patrilineality or matrilineality of kinship systems. The textbook says
so too. So did at least one video I showed to the class.


In our society, there has been a recent increase in the number of
single-parent households headed by women. This is associated with an
increase in the A)(41%)number of children below the poverty line;
B)(33%)participation of extended kin in caring for the children;
C(10%)amount of child support being paid by fathers; D)(16%) effective
income because of welfare and reduced housing costs.

Well, more people picked the right answer than any other. But the
discrimination index was NEGATIVE: the people who picked the right answer
were the wrong people!


Of course, there were a few questions that didn't fit the pattern I think
I see in the cluster of anomalous ones. I haven't figured out why it
didn't work, but here's one that got a discrimination index of just about

On average, students in this class are more likely to marry someone who
is now a student at N.I.U. than a member of any other class of people.
This is a good illustration of the meaning of
A)(2%) incest; B)(47%) endogamy; C)(44%) exogamy; D)(7%) polygyny.

Two weeks before the test I distributed a list of words and told them if
they got the meanings of those words pretty clear in their heads they
would do pretty well on the next test. All of these words were on the
list. (So were matriarchy and patriarchy, by the way.) I deliberately
did not provide the definitions on that piece of paper. I told them why:
I thought there was more chance that they would remember the definitions
if they had to look them up themselves. (But I did define every one of
them, in passing, during one lecture or another during this section of
the course.)


Now I'm back to my "what's going on here?" question. There's something
quite striking about the unusually high number of questions that all
dealt with essentially the same subject, relations between the sexes,
and turned out to have disturbing statistics. I've never seen anything
similar on the results of my tests. I certainly didn't expect the result
when so many of the questions that performed strangely were pre-tested
and had not given such marked problems before.

(There were several more questions that showed the same pattern, but I
got tired of typing and figured you were probably going to be tired of
reading any more of them.)

Maybe Dan Foss's question about "nuclear family" is another facet of
whatever I've tapped into here.

Anybody care to pontificate on the significance of this all?

--mike salovesh, anthropology department <salovesh@niu.edu>
northern illinois university PEACE !