chabot@amber.DEC (Lisa Chabot) (05/24/84)
I quote a statement from Mr. Hughes: > Path: decwrl!decvax!hoopoe!aliases!burl!cloud!watmath!csc > Subject: Re: Tests and Prediction > Posted: Sun May 20 10:54:59 1984 > > One often hears the statements "Tests can only measure the ability to take > tests" and "Tests can only predict the ability to take tests". The first > is tautological (in terms of the TQ analogy, A test of Tallness can only > measure a persons tallness) the second is simply false. This is an incorrect fitting to the "TQ analogy". A correct interpretation would be A test of Tallness can only measure a person's ability to take tallness tests. In this correct fitting to the "TQ analogy", it is readily apparent that this is not a tautology. Those who cannot recognize the difference perhaps are deficient in abstract reasoning, but this may be due to environmental factors such as the lack of good practice in verbal reasoning in many engineering environments, rather than any innate lack. To assist in visualization about the difference between "a person's ability to take tallness tests" and "a person's tallness", consider the following: we wish to test a sufficient number of male and female subjects but our subjects come from a country or societal structure in which women are not allowed out of the home to protect them from interactions with men not of their household , or in which women are merely wary of such interactions. We, the researchers may either be ignorant of such a wariness or unable to overcome any such prohibition if we are not welcome into potential samples' homes for reasons of the male researchers among us not being welcome or our Western taint not being welcome. In this situation, we may not be able to reach the same variety of women that we can of men, certainly any women who approached us and we were able to measure might be suspected of being abnormal, at least by the standards of their society and their freedom to approach might result from some tolerance because of a physical feature decided by their society to be a defect. Being unaware of the outlying factors determining what sample we pick can grossly distort our research. There was a study done on (I believe) Norwegian women which became the basis for the widely held belief that women used to reach menarche at age 18; this contrasted with what we see around us today to be the more common age of 13 led many to wonder or fear just what was happening to the human race, and if this was perhaps some sign of moral decay, or mutation, or a result of overcrowding. But it was simply this: the women studied were starving women, and it has since been shown that the onset of menstruation is delayed if a woman's proportion of fat in body weight is too low. Statements of the form "Tests can only measure/predict an ability to take tests" are perhaps incorrect after all in that the "only" is too restrictive. "Tests measure/predict the ability to take tests" is more correct. "Tests measure the researchers' ability to design tests" is most correct. --Lisa S. Chabot UUCP: ...{ decvax | allegra | ucbvax }!decwrl!rhea!amber!chabot USFail: DEC, MR03-1/K20, 2 Iron Way, Marlborough, MA 01752
martillo@ihuxt.UUCP (Yehoyaqim Martillo) (05/24/84)
Ability to solve a certain type of math problem within a specified amount
of time is inherently much more testable or measurable than a less well
defined quality like I.Q.
Lisa Chabot and many others claim that math anxiety is inculcated in women
in American society. Such inculcation may be true, but I suspect all
Americans suffer from Math fear and there is no gender specificity
whatsoever. In France, linear algebra is taught at high school level. In
Austria calculus begins to be taught at 2nd year of the Gymnasium.
Americans at all levels are well behind Europeans and Japanese in math.
--
Yehoyaqim Shemtob Martillo
(An Equal Opportunity Offender)csc@watmath.UUCP (Computer Sci Club) (05/24/84)
Lisa Chabot attempts to find a difference between test score and the
ability to take a test, by showing an example where a population
mean is incorrectly estimated due to a poor choice of sample!
Does Lisa Chabot suggest that the age of mencarche of the women
studied was incorrectly determined because they were not from
a representative sample?
I agree that there is a big difference between saying that
the sample of women (men, monkeys, wombats) we tested had a mean
of x, and saying the mean of all women is x. Also even if
sampling technique is good, results of a survey can only apply
to the population from which the survey was drawn. (Good sampling
techniqe and population are related here. Good sampling technique
means that each member of the population has an equal chance of
being selected) Extrapolations to other populations are dangerous
(even though they are quite often valid).
There is a difference beween the term I.Q. as it is sometimes used
and the score on some specific I.Q. test. That is because research
has led us to believe there is some specific set of attributes of
human beings which we can call I.Q. This I.Q. is highly correlated
to scores on certain tests, but not identical. That is one might
raise ones score on a test by training in the methods of this test
yet not be said to have raised ones I.Q. When I talk about I.Q.
I mean the score on some specific I.Q. test.
What does all this have to do with saying there is a difference
between a persons tallness and their ability to take tallness tests?
William Hughes