gazit@cs.duke.EDU (Hillel Gazit) (01/18/91)
In article <IbZPvNe00WB9IEwqJE@andrew.cmu.edu> ag1v+@andrew.cmu.EDU ("Andrea B. Gansley-Ortiz") writes: >A couple years ago a black man redid the standard IQ test in such a >way that it centered on a black community's knowledge. The results >were that blacks scored above average while whites/other minorities >did not score as well. I have no doubt that it can be done, the only question is what do you suggest to do about that; there are two choices: 1) To skew the exams in the direction of the group *you* like. 2) To throw all these exams out of the window. Before you answer, I want to tell you a story: When I was accepted to the Technion, Israel they used entrance exams in math and physics. The idea was that if your grades were higher than the matriculation grades then they did not care about the matriculation; if they were lower then they average the matriculation and the exam grade. No IQ tests. I asked, informally, what was the reason and the answer I got, informally, was that they checked the idea and the tests in math and physics were a better prediction *for their students*. (When they ignored the matriculation grade they got a better prediction, but they could not continue the practice because of political pressures.) The result of the Technion's system was that people who did not do too well in high school had a second chance - pass these two hard exams. The results were that: 1) Students who came from the best high schools with excellent matriculation suddenly had to compete with people like me with below average matriculation. It was nice to see an over-inflated ego lose air... 2) The Technion picked some of the best students (e.g. I and Salit) who had no chance to be accepted to other universities. Good students give an edge to a university, and it helped to keep the Technion as the best technological institute in Israel. 3) Comparing the results of the entrance exams with the matriculation proved that something was (and is) rotten in the matriculation system. Later Political pressure forced the Technion to adopt a system of IQ entrance exams similar to the other universities, and give up their method. The free market responded by expensive courses that can raise the IQ grade by 30 points in a couple of months. (I don't think that any course could do something similar to math and physics grades...) I hope that you can see why I think that IQ tests are not the right tool to sort applicants, and that's why people who want to discriminate are so happy to use them... >ag Hillel gazit@cs.duke.edu "...13 of 17 valedictorians in Boston high schools last spring were immigrants or children of immigrants." -- US. News & World Report, May 14, 1990
milt%odin.corp.sgi.com@sgi.COM (Milton Tinkoff) (01/19/91)
In article <664152739@grad17.cs.duke.edu> gazit@cs.duke.EDU (Hillel Gazit) writes: >In article <IbZPvNe00WB9IEwqJE@andrew.cmu.edu> ag1v+@andrew.cmu.EDU ("Andrea B. Gansley-Ortiz") writes: > >>A couple years ago a black man redid the standard IQ test in such a >>way that it centered on a black community's knowledge. The results >>were that blacks scored above average while whites/other minorities >>did not score as well. I've _heard_ of this but I'm a bit confused. Exactly how is an IQ test racially biased? I thought typical questions were of the "What is the next number in this sequence?" or "Square is to circle as cube is to ?" variety. Show me some racially biased questions and then I'll start believing. -- ------------------------------------------------------------------------------- Milt Tinkoff | "The average man is a Silicon Graphics Inc. | stupid man." milt@waynes-world.esd.sgi.com | -Ed Mao
RMG3@psuvm.psu.EDU (01/25/91)
In article <1991Jan21.014257.16464@nntp-server.caltech.edu>, hmj@miranda.caltech.EDU (Helen Johnston) says: >In article <1991Jan18.174824.21081@odin.corp.sgi.com> >milt%odin.corp.sgi.com@sgi.COM (Milton Tinkoff) writes: >> >>I've _heard_ of this but I'm a bit confused. Exactly how is an IQ >>test racially biased? I thought typical questions were of the "What >>is the next number in this sequence?" or "Square is to circle as cube >>is to ?" variety. Show me some racially biased questions and then >>I'll start believing. >I don't know about racially biased, but culturally biased is easy. I >am Australian, and so had barely heard of baseball, let alone knowing >any rules. [GRE example] My mother encountered a problem on the SAT which asked (roughly) 'How much grass in fair territory in the infield could a goat on a 30 foot tether eat?'. Anyone without detailed knowledge of the layout of a baseball field (such as my mother) has no hope of answering this one. A much better place to look, rather than relying on anecdotes is Stephen Jay Gould's book 'The Mismeasure of Man'. In it he details the history of mental testing and the abuses thereof. It includes the biases built in against people not like the people who write the tests (economically, socially, by gender, religion, ...) Some crucial points from the book, taking IQ as the example: 1) You have to assume that there is such a thing as intelligence. 2) You _assume_ that it is measureable. 3) You _assume_ that it can further be represented by a _single_ number. 4) You _assume_ that your method of measurement has no effect on the result. Another piece about IQ's. The raw score on standard IQ tests levels out after about age 20, so that an average 20 year old and an average 40 year old (whatever _that_ means) score the same. Now, IQ refers to dividing the 'mental' age by the calendar age. So all 20 year olds are geniuses (having a mental age of 40 and a calendar age of 20) and all 40 years olds are morons (having a mental age of 20 and a calendar age of 40). Substitute the groups of your choice for the 20 and 40 year olds in my example and you have the arguments advanced for 'proving' that group xyz is less intelligent than group pdq. Bob Grumbine Osgood's law: Variables don't and constants aren't.
zvs@bby.oz.au (Zev Sero) (01/25/91)
Helen = hmj@miranda.caltech.EDU (Helen Johnston) Helen> When I took the GRE to come to school here, I remember one Helen> of the logic questions - one of those tests they are so fond of, that Helen> have a whole set of conditions (A hits before C does, B hits two Helen> places after F etc.) which you then have to come up with a solution Helen> for. They gave a set of rules that included things like "Once a Helen> pitcher has been removed from the lineup, he cannot pitch again". I Helen> dutifully read through all these rules several times and solved the Helen> problem. The point was, of course, that any American, and in Helen> particular any American male, would have only needed to glance at the Helen> rules to check whether they are the right ones. I of course was Helen> treating them as abstract conditions, and so presumably took much Helen> longer on the problem. Perhaps the rules given were deliberately not the `right' ones, and that by automatically treating them as abstractions you were actually at an advantage over a baseball fan who would misread the rules or not read them at all, and get the wrong answer. Such `trick' questions are fairly common in that type of test, since what they are trying to measure is your ability to understand and correctly apply information that you are given. -- Zev Sero - zvs@bby.oz.au This I say unto you, be not sexist pigs. - The prophetess, Morgori Oestrydingh (S.Tepper)
marie%singsing.Berkeley.EDU@ucbvax.berkeley.EDU (Marie desJardins) (01/29/91)
In article <1991Jan18.174824.21081@odin.corp.sgi.com> milt%odin.corp.sgi.com@sgi.COM (Milton Tinkoff) writes: > >I've _heard_ of this but I'm a bit confused. Exactly how is an IQ >test racially biased? I thought typical questions were of the "What >is the next number in this sequence?" or "Square is to circle as cube >is to ?" variety. Show me some racially biased questions and then >I'll start believing. I recently saw some IQ tests that had a number of questions where you were supposed to unscramble five sets of letters, and indicate which of them did not belong to a particular group. A lot of these questions relied on knowledge one would learn in school or in "common" (middle-class white) experience, e.g. European cities (some of which I had never heard of!), automobile makes, etc. These are certainly biased in the sense that they measure acquired, not innate knowledge, and that the acquired knowledge measured is more likely to be known by certain groups (in some cases males, in some cases well-educated or well-travelled people---usually whites). These were fairly old tests (late '60's, maybe), but I imagine similar questions are still found on current tests (in fact, I wouldn't be surprised if they still use the same tests!) Marie desJardins marie@singsing.berkeley.edu
ellene@microsoft.UUCP (Ellen Mitsue Eades) (01/31/91)
Reply-To:ellene@microsof.uucp (Ellen Eades) In article <1991Jan21.014257.16464@nntp-server.caltech.edu> Helen Johnston <miranda!hmj@tybalt.caltech.EDU> writes: >The New York Times last year gave a few more examples that people >complain about. Things like, in the verbal section, "A stockbroker is >to a [...] as a [...] is to what?" I distinctly remember one question on the verbal section of the SAT in which the example, "A bull is to a bear as a [...] is to a [...]." If you aren't familiar with the stock market you'll certainly never encounter these terms in proximity. And there aren't that many inner- city schools that teach economics before your senior year, if at all. Ellen Eades
zvs@bby.oz.au (Zev Sero) (02/27/91)
Bob = RMG3@psuvm.psu.EDU (Bob Grumbine) Bob> My mother encountered a problem on the SAT which asked (roughly) Bob> 'How much grass in fair territory in the infield could a goat on Bob> a 30 foot tether eat?'. Anyone without detailed knowledge of the Bob> layout of a baseball field (such as my mother) has no hope of Bob> answering this one. If presented with that question, I would assume that the `in fair territory in the infield' business was irrelevant misdirection, and the real question was `what is the area of a circle with a radius of 30 feet?', so I would answer `as much grass as covers 9000 x pi square feet'. It's the sort of thing I would expect from that sort of test, to weed out the ones who dredge up their baseball knowledge instead of concentrating on the real question. Could some kind American tell me if I am right? Is the fair territory in the infield (whatever that is) greater than 60 feet across? BTW, if the question really does depend on these details, then the question becomes woefully deficient, because it doesn't specify that the tether is placed anywhere near a baseball field; even if it is, it might be right on the boundary of the said area, so that only half the circle is inside the area; it is not specified how long the goat is left there---not long enough to eat all the grass in reach? long enough for the grass to regrow and be eaten again?---or how big the goat is, etc. I just had a horrible thought---is there any grass at all in the fair territory in the infield? It wouldn't be a concrete area, or one mown and rolled down to nothing, would it? -- Zev Sero - zvs@bby.oz.au This I say unto you, be not sexist pigs. - The prophetess, Morgori Oestrydingh (S. Tepper)
sderby@bcm.tmc.edu (Stuart P. Derby) (02/27/91)
> [Previous poster asks how IQ tests can be racially biased]
Well, to be precise, the questions are not "racially biased", but
rather culturally biased in a way that often correlates strongly with
race, at least at some place and time. As an example, a drawing of two
people playing tennis, with rackets omitted, is presented, and the
testee is asked to select the missing item from a set of drawings. Or
similarly, a table, set with 4 tea cups, sans saucers, is presented,
and the obvious question asked. Since these questions actually
measure cultural literacy (at least, in an untimed exam - speed might
be considered a measurement of intelligence, if set up correctly) with
a white, American/North European bias, I would loosely term the
questions racist.
If memory serves correctly, the above examples were snitched from
Stephen Jay Gould's excellent book _The Mismeasure of Man_, where I
believe they were cited as actual questions used in the intelligence
exams of prospective immigrants to the U.S., in the 1920's or so. The
"evidence" thus accumulated was used as an argument within Congress in
debating immigration restrictions. The subsequent immigration law
favored north Europeans by setting immigration quotas to reflect the
ratios of country-of-origin of the U.S. population as of 1890 (or so).
_The Mismeasure of Man_ discusses racial issues in fair depth,
including other flawed measures of intelligence. Readers of this group
will also be interested in a story of forced sterilization here in the
good ol' U.S. of A, as approved by the Supreme Court, the Carrie Buck
case.
-Stucdm@inel.gov (Dale Cook) (03/01/91)
In article <1991Jan28.234342.10420@melba.bby.oz.au>, zvs@bby.oz.au (Zev Sero) writes: |> Bob = RMG3@psuvm.psu.EDU (Bob Grumbine) |> |> Bob> My mother encountered a problem on the SAT which asked (roughly) |> Bob> 'How much grass in fair territory in the infield could a goat on |> Bob> a 30 foot tether eat?'. Anyone without detailed knowledge of the |> Bob> layout of a baseball field (such as my mother) has no hope of |> Bob> answering this one. |> |> I just had a horrible thought---is there any grass at all in the fair |> territory in the infield? It wouldn't be a concrete area, or one mown |> and rolled down to nothing, would it? |> No, but it could be astroturf :-) :-) :-) (Sorry, couldn't resist...) ---------------------------------------------------------------------- --- Dale Cook cdm@inel.gov The opinions are mine. The following disclaimer is my employers. ---------------------------------------------------------------------- ========== long legal disclaimer follows, press n to skip =========== Neither the United States Government or the Idaho National Engineering Laboratory or any of their employees, makes any warranty, whatsoever, implied, or assumes any legal liability or responsibility regarding any information, disclosed, or represents that its use would not infringe privately owned rights. No specific reference constitutes or implies endorsement, recommendation, or favoring by the United States Government or the Idaho National Engineering Laboratory. The views and opinions expressed herein do not necessarily reflect those of the United States Government or the Idaho National Engineering Laboratory, and shall not be used for advertising or product endorsement purposes.