psych504.htm

Psychology 504

Schonemann

Fall Semesters

3 credit hours

Prerequisites: some background in elementary statistics (e.g., psych 500) or consent of instructor

Grade: active class participation and term papers

1. Fancher: The Intelligence Men.  New York: Norton, 1985

2. Glauberman: The Bell Curve Debate. New York: Time Books, 1995.

3. Owen: None Of The Above. Boston: Houghton Mifflin. Presently out of print. Copies on loan from the instructor.

Virtually since the inception of the mental test movement in the 20s, controversy has surrounded IQ tests. After some  relative quiescence in the 50s and 60s, this debate  flared up again in the late 60s, when Arthur Jensen, an educational psychologist at Berkeley, published a provocative paper entitled: How much can we boost IQ and scholastic achievement? in Harvard Educational Review.

In a nutshell, his basic theses were:

(a) "Intelligence" is operationally definable as Spearman's g and measureable by IQ tests.

(b) This intelligence is major determinant of academic, occupational , and social success .

(c)  this intelligence is substantially  inherited  (the initial figure was 80%).

(d) some ethnic groups are, on average, inferior to other groups on this  trait.

More recently, Herrnstein and Murray (1994) have reititerated these theses in a widely publicized book, The Bell Curve and discussed some of the social implications, as they saw them.

The purpose of this course is to review the evidence  behind these claims and to lay the ground for a rational discussion of them.

A central theme of the course is the question of definition: What, exactly, is "intelligence"?  While numerous definitions have been proposed  over the course of the last century, the experts still cannot agree on what this term is supposed to mean.  This raises the question how they can possibly know whether a particular IQ test "measures" intelligence, and how they  found out how much of it is inherited. These are some of question we will discuss in this course.

One premise in designing it was that anybody with an open mind is as qualified as the experts to discuss these questions. Another premise is that these questions should be discussed, because IQ tests and there derivatives are still widely used in the US, and  some of the answers of the experts are fraught with severe adverse consequenes for  many people if they should turnout to be wrong. In fact, as will be shown, they already did have harmful consequences, so that it would be socially irresponsible to pretend the IQ Controversy is of purely academic interest.

An open mind and some willingness to engage in independent study.

While some of the technical discussions do involve elementary statistical notions (such as means, variances, correlations, probabilities, and derived concepts such as factors,  principal components, and conditional probabilities), the literature shows that most  modern test experts are quite cavalier in their use of these technical terms. Hence it would be unfair to expect more from students. Thus, while some prior exposure to elementary statistics  would not hurt, it is not necessary for openminded students to benefit from this course, especially since some of the needed concepts will be briefly reviewed at the early stage. Students without statistical background might in fact benefit from the course as a preview of statistics in action.

1.  History of  IQ, pre-US

Overview. Chinese civil service system. Mediaeval Europe. Ebbinghaus. Wundt vs. Galton.

2. Galton: normal curve, correlation, regression towards the mean, twin method, Heriditary Genius. Wundt, McK. Cattell, Wissler. Binet, Simon, Stern, "IQ".
 
3. History of IQ, US

Goddard, Kallikaks, definition of "feebleminded". Immigration problem, "nordics", Jews, "Yellow Peril". Immigration and sterilization laws. Terman's Stanford Binet test, longitudinal study of "gifted" children. WW1: Army tests. Brigham: "The American Intelligence", scholastic aptitude tests.

4. The Wechsler series. Factor analytic rationale (roughly): Spearman, Thompson, Thurstone, R.B. Cattell, Jensen. Dissenting voices: Lippman, Piaget, Sternberg.
 
5. Statistical Interlude: Test construction and evaluation

Review of elementary statistical concepts: mean, variance, correlation. Item analysis. Reliability, Spearman-Brown prophecy formula , "Cronbach's" alpha. Validities: types of,  correction for attenuation, correction for restriction of range, Taylor-Russel tables.

6. A closer look at factor analysis. Classical True Score Theory as a special case. How Principal Components differ from factors.  Number of  factors problem, factor indeterminacy, lack of fit.
 
7. Commercial IQ tests and their predictive validities

Binet tests: Proverbs. (WWI) Army tests.

8.. Wechsler series. Thurstone's PMA. "Culure-free" IQ tests, esp. Raven's Progressive Matrices.
 
9. Scholastic Aptitude Tests

 Relation to IQ tests. ETS. SAT, GRE.  Coaching. Reading comprehension (without reading).

10. Crouse-Trusheim validity analysis of the SAT. Humphreys "fleeting validities".

11. Heritability

Intraclass correlations, variance component models, "heritability estimates". Nature/nurture problem.

12. Twin and adoption studies. MZT/MZA data. The Burt scandal. Jinks and Fulker. DZT/MZT data. Main problem: The data violate the assumptions of the models.

13. "Race", also a problem of definition. Galton, Shuey, Jensen, Rushton. Eyfert's GI data.  Are"Orientals"  genetically superior? Lynn and Flynn.

14. Odds and ends as time allows:

Constancy of IQ question. Cattell's prediction, Flynn's results. Hit-rate bias. NCAA admission standards controversy, Spearman Hypothesis.

15. Remainder: term paper presentations.