The Religious A priori
An evaluation of Atheist Claims
Who is Smarter?page 3
Even if we were to grant the studies themselves, the over all argument is still beging the question. Who says that IQ tests and SAT scores are relly indicative of intelligence? Some of the studies used grades as measurement, but many of them used SAT scores. This is inadeqate, as Standardized testing is inadequate as a measure of intelligence.
I. Standardized tests do not measure intelligence, or academic performance.
Race, adacemic Achievement, and School Reform
An analysis of the racial consequences of state mandted testing
Harold Berlak holds a doctorate educational reseach from Harvard. He is an former professor of education at Wshinton University in St. Louis, and lontime educational activist.
Numerous researchers have carefully documented the highly disproportionate adverse impact on students of color of standardized achievement testing. (16) An argument might be made that these differences in test score while small, nevertheless represent real differences in performance, and that tests though imperfect eliminate the incompetent, those most likely to perform poorly at school or on the job. Steele's study suggests the opposite --that the more talented students are at greater risk of failure. As documented in section 4, there is no evidence to support the claim that standardized tests are valid and credible measures of academic achievement or intellectual capacity. There is no demonstrable connection between observed academic performance and standardized test scores. Test scores do not predict future success in school, the university, or in the workplace. In the case of CBEST, several studies were conducted to explore the link. CBEST showed no correlation to current or future performance on the job, or to observed levels of literacy. While some tests do correlate statistically to future grades, this correlation is short lived. (17) What standardized achievement tests appear to predict best are scores on other similarly constructed tests, and parent's wealth. As reported by Peter Sacks, socio-economic class accounts for approximately 50% of the variance in SAT test scores. He estimates that for every additional $10,000 in family income, a person on average gains 30 points on the SAT.(18)
The most fundamental problem with both types of standardized academic achievement tests is that there is little evidence to support the contention that they measure what they purport to measure -academic achievement, or proficiency. This does not mean that academic achievement and high standards are not vital. Rather it is that the tests have very little relationship to actual academic performance of any kind. For some standardized tests there is correlation to grades at least in the short term. But, for virtually all standardized reading and writing tests, there is no demonstrable connection between a person's performance on a standardized reading test and a person's reading abilities in the real world --in everyday life situations, at school, work and elsewhere where one might be called on to read. What this means is that contrary to common sense, Student A's , score at the 45th percentile and Student B's at 95th percentile on the Stat-9 reading test (or any other norm-referenced tests) says nothing whatsoever about the actual or relative reading performance of students A or B. The standardized test informs us only how every test-taker's score on the test compares to everyone else taking the same test.
A score on basic skills' or 'proficiency' test tells us only how far above or below the established cut-off a student's score falls. Cut scores on academic proficiency test are not based on actual or observed level of competence or proficiency. There have been numerous studies that have explored the relationship between test performance and actual performance, and researchers have repeatedly come up with the same conclusion: no (or almost no) connections. Neither do the tests meet the criterion of 'predictive validity'. Norm-referenced and proficiency tests (except for grades in the short term) do not predict future success in school, the university, or in the workplace. (21) What the tests predict best is a person's score on similarly constructed test and parent wealth.
Intelligence and Mental Tests
J. R. Kantor
University of Chicago
And now we must consider what light the work on psychological tests throws upon the problem of intelligence. A study of the actual procedure and results of mental tests proves conclusively that such tests are and can only be designed to measure some performance whose achievement is the result of a previous interaction of a person and objects (machines----materials). It is for this reason that "no test has any significance for employment purposes until it has been tried out on employees doing exactly the same kind of work as that for which new applicants are to be tested later on.10 Illuminating in the extreme in this connection is the study of the limitations of mental tests. What must one conclude from the fact that mental tests are of no service in selecting executives? Should we say that mental tests do not attempt to measure intelligence? For surely, if they did, they could not be applied to any more directly functioning intelligence than is found in the work of an executive. But to accept this conclusion would mean giving up the whole problem of measuring intelligence, and this is impossible, for the genuine usefulness of the tests indicates that there may be degrees of intelligence, the lower ones of which may be very readily determined. Or should we say that intelligence is an unknowable thing, at least so far as tests are concerned, since tests are only useful for acts which have a definitely standardized form? To the writer it seems that the difficulty is entirely factitious and based upon the misconception that intelligence is native.What the inapplicability of tests to the selection of executives really teaches us is, that all tests are performance tests based upon definite reaction patterns and not measures of connate capacity.
Now since executive intelligence means the possession of innumerable and complex reaction systems it is entirely to be expected that the present development of tests should be still inadequate to meet the situation. And, further, the student of tests must be always unable to meet this situation if he persists in the belief that intelligence is innate, since such a view precludes the investigation of the actual contributing conditions which make possible complex human adjustments. To mention just one difficulty, the applied psychologist makes too wide a difference between moral and mental qualities, as though it were possible completely to separate these when an employment problem is under investigation. In this connection it is remarkable to observe upon what slender threads are sometimes hung the belief in an absolute intelligence factor. Thus the positive correlation between tapping, letter crossing, and other tests is presumed to be evidence of the presence of such a general intelligence factor.
The National Center for Fair & Open Testing
342 Broadway Cambridge, MA 02139 (617)864-4810 (617) 497-2224 (fax)
Website: www.FairTest.org, e-mail: FairTest@aol.com
Do test scores reflect real differences among people?
Not necessarily. To construct a norm-referenced test (a test on which half the test-takers score above average, the other half below), test makers must make small differences among people appear large. Because item content differs from one test to another, even tests that claim to measure the same thing often produce very different results. Because of measurement error, two people with very different scores on one test administration might get the same scores on a second administration. On the SAT, for example, the test-makers admit that two students' scores must differ by at least 144 points (out of 1600) before they are willing to say the students' measured abilities really differ.
Do IQ tests measure intelligence?
IQ tests assume that intelligence is one thing that can be easily measured and put on a scale, rather than a variety of abilities. They also assume intelligence is fixed and permanent. However, psychologists cannot agree whether there is one thing that can be called intelligence, or whether it is fixed, let alone meaningfully measure "it." Studies have shown that IQ scores can be changed by training, nutrition, or simply by having more friendly people administer the test. In reality, IQ tests are nothing more than a type of achievement test which primarily measures knowledge of standard English and exposure to the cultural experiences of middle class whites.
Next: Page 4