How changeable are IQ test scores?

1. Introduction
A continuation of the critical thinking theme of interpreting scientific ideas in relation to the actions favored by the exponents of those ideas is to pose a question that can be asked of all sides: What can we do (on the basis of the science)?

1b. Mini-lecture (interactive)

2. Reading
A famous debate occurred around 1970 between Jensen (1923-2012) an educational psychometrician (analyst of psychological data) who believed that IQ was inherited and difficult to change and Lewontin (b. 1929), a population geneticist (evolutionary biologist) who argued that Jensen's method was flawed and who believed that society had not tried very hard to boost intelligence. The reading does not include the original Jensen article in the Harvard Educational Review because it is too long, so you have to discern his arguments from Lewontin's account of them and Jensen's response to Lewontin. Read and unpack their exchange by making up a charts along the following lines, where you add rows for each new point, leave a cell blank if there is no response by the other person, and use "Notes" to record your response to their points, e.g. What convinces you? What don't you understand? (It is difficult material so don't be afraid to say when you don't follow the technicalities.) When you look at the exchange as a whole, consider what political, social or other assumptions are involved in the argument. What actually gets answered and what not?

Jensen's arguments in Lewontin's account
 Lewontin on specific points of Jensen's arguments
 Notes
e.g., because blacks perform, on average, more poorly than whites on IQ tests, and because compensatory education programs have failed to remove this difference (due to his hypothesis that IQ is genetic), there is no use trying to remove the difference in IQ through education.
 ...................
 ...................


...................

Jensen's (re)view of his own arguments
 Notes
e.g., It is necessary to look at individual differences rather than taking it as read that “all children are alike.” This could be more fruitful than large scale programs based on a philosophy of general cultural enrichment.
 ...................

Jensen on the points of Lewontin's account
 Lewontin's response to Jensen on his (Lewontin's) account
 Notes
e.g., Lewontin's paper has an ad hominem flavor.
...................

Additional questions to prepare for class discussion: Jensen asserts that equal education does not work and that we need education tailored to the individual needs of students. Does he give any evidence that equal education had been attempted in the USA by the late 1960s? Do you know of any such evidence? Who do you think would find his assertion plausible (c. 1970) with little or no evidence given? What is the context that makes this plausible? If you need help thinking about these questions, do some internet investigation.

3. Activity
Review the charts of arguments, counter-arguments, and missing arguments in the exchange
What can we do on the basis of the science of Jensen? Of Lewontin?
Review how this relates to the critical thinking theme of interpreting scientific ideas in relation to the actions favored by the exponents of those ideas.


4. Synthesis and extensions

By unpacking the debate between Lewontin and Jensen, the following themes should emerge:
1. There are no simple explanation about genetics and socially significant traits, especially regarding average racial differences. Therefore: be skeptical of anyone who proposes such an explanation (i.e., scrutinize where they are coming from).
2. There are researchers who have detailed technical cases to make (even if their conclusion is quite simple, e.g., Jensen believes it is plausible that genes account for (average) racial differences). Their arguments can be teased out into their components (in this case, with help from Lewontin, who can handle the technical side and has views about the political/social implications and underpinnings of the science).
3. What can we do (on the basis of the science) is worth asking of all sides. (Jensen thinks we have tried and failed to equalize education so educate people according to their innate capacities [actually, educate people according to the typological generalization about the average differences between the races]; Lewontin thinks educational professionals have not tried very hard and when society is committed to equalizing education ways will be found.)

Lewontin vs. Jensen debate

A famous debate c. 1970 between an educational theorist (Jensen) who believed that IQ was inherited and difficult to change and a geneticist (Lewontin) who argued that Jensen's method was flawed and who believed that society hadn't tried very hard to boost intelligence (pjt).

Precis (by JanCoe) of Lewontin, R. (1976). “Race and Intelligence (and Jensen's reply, and Lewontin's reply to that).” In The IQ Controversy: Critical Readings, ed. N. J. Block and A. Dworkin. NY: Pantheon, 78-112

Jensen's arguments in Lewontin's analysis
 Lewontin on specific points of Jensen's arguments
Jensen's overall argument is that: because blacks perform, on average, more poorly than whites on IQ tests, and because compensatory education programs have failed to remove this difference (due to his hypothesis that IQ is genetic), there is no use trying to remove the difference in IQ through education.
Since compensatory education has failed, shouldn’t we inquire why that is so?
 Jensen's professionalist bias that makes him take the position that if a problem were soluble, it would have been solved.
IQ tests are culture bound and there is good reason they should be because they are predictors of culture-bound activities and values and correlate highly with occupational status.
 Jensen is not arguing that that IQ is environmentally determined because it is culture bound. He is saying that IQ is culture bound in the sense that it is related to performance in a Western industrial society. The determination of the ability to perform a culturally defined task might itself be entirely genetic.
If the poorer performance by blacks on IQ tests has largely genetic rather than environmental causes, it follows that blacks are also genetically handicapped for other high status components of Western culture.
 Jensen has seen that this argument cuts both ways.
Therefore, IQ testing is simply ONE manifestation of these genetically determined differences.
In discussing the causes of the differences in IQ results between blacks and whites, two issues arise: 1) the stability of the IQ through an individual's lifetime; 2) the genetic basis of IQ.
“The literature” indicates that IQ is more or less set by the age of 8.
 To say that children do not change their IQ is not the same as saying that they cannot.
The genetic argument deals with the distribution and inheritance of intelligence. Jensen demonstrates that IQ scores are normally distributed.
 There is nothing in genetic theory that requires, or even suggests, that a phenotypic character should be normally distributed.
The “underlying normality” of the distribution appears as a consequence of the genetic control of IQ.
 We cannot speak of a trait being molded by heredity, as opposed to environment.
Every character of an organism is the result of a unique interaction between the inherited genetic information and the sequence of environments through which the organism has passed during its development.
Heritability, its application to a specific population in a specific set of environments, and the difficulties in its accurate estimation are all discussed by Jensen – he does grasp the technical issues involved.
 The heritability of a measurement is defined as the ratio of the variance due to the differences between the genotypes to the total variance in the populations. Different populations may have more or less genetic variation for the same character.
Concerning the causes of the difference between the IQ distribution of blacks and whites, Jensen concludes that genetic factors are strongly implicated.
 The evidence that Jensen has offered is IRRELEVANT to the question. Jensen confuses the heritability of a character within a population with the heritability of the difference between two populations.
The genetic basis of the difference between two populations bears no logical or empirical relation to the heritability between populations and cannot be inferred from it.
Jensen's second conclusion is that compensatory education for blacks has failed.
 It is empirically wrong to argue that, if the richest environment experience we can conceive does not raise IQ substantially, that we have exhausted the environmental possibilities. The supposition that compensatory education must fail arises from a misapprehension about the fixity of genetic traits.
There is no reason to believe that the IQs of deprived children, given an environment of abundance, would rise to a higher level than the already privileged children's IQs.
 On the reasonable assumption that ways of significantly altering mental capacities can be developed if it is important enough to do so, the real issue is what the goals of our society will be.

Jensen's (re)view of his own arguments
Based on the massive evidence presented by the Civil Rights Commission that compensatory programs have produced no significant improvement in measured intelligence or scholastic performance of disadvantaged children, merely to apply more of the same will not likely lead to the desired results.
It is necessary to look at individual differences rather than taking it as read that “all children are alike.” This could be more fruitful than large scale programs based on a philosophy of general cultural enrichment.
His theory is that there are two broad categories of mental abilities: intelligence and associative leaning ability. Large racial and social differences are found for intelligence but negligible differences for associative learning abilities.
IQ ability is a selection of just one portion of the spectrum of human mental abilities, but is important to our society
The methods and evidence he reviewed in his paper led him to his conclusion that individual differences in intelligence are predominately attributable to genetic differences. He notes the lack of heritability studies in the minority populations and says they are needed to increase understanding of what our tests measure in those populations.
IQ differs, on average, among children from different social class backgrounds and the evidence indicates to him that some of this is attributable to environmental differences and some to genetic differences among social classes.
Social scientists decree on purely ideological grounds that all races are identical in the genetic factors that condition various traits, including intelligence.
But nearly every anatomical, physiological, and biochemical system investigated shows racial differences. Why should the brain be any exception? A genetic hypothesis is not unwarranted.
A problem that is more socially important than the question of racial differences per se, is the dysgenic trends that can be observed. That is, if the poorest blacks keep having the most children, then the genetic intelligence difference between whites and backs could widen even further.

Jensen on the points of Lewontin's analysis
 Lewontin's response to Jensen on his (Lewontin's) analysis
Jensen's article is not an objective empirical scientific paper that stands or falls on the correctness of his calculation of heritability. It is a closely-reasoned ideological document springing from a professionalist bias and permeated with an elitist and competitive worldview.
Lewontin's paper has an ad hominem flavor.
 There was no ad hominem argument in Lewontin's analysis.
He agrees with Lewontin that the assumptions, theories and practices of educators, social and behavioral scientists are bankrupt.
Re: Lewontin's remark “to say that children do not change their IQ is not the same as saying that they cannot” – Jensen has never said anything to the contrary.
Re: Lewontin's argument that heritability of a trait within a population does not prove that genetic factors are involved in the mean difference between two different populations on the same trait – is one that Jensen never advocated. It is a straw man argument set up by Lewontin.
Theoretically, Lewontin's statement that heritability within populations is irrelevant to the question of genetic differences between differences is true; however, it is necessary to distinguish between the possible and the probable. The real question is not whether a heritability estimate, by its mathematical logic, can prove the existence of a genetic difference between two group, but whether there is any probabilistic connection between the magnitude of the heritability an the magnitude of the group differences.
 Jensen has responded to Lewontin's major scientific thrust at his thesis by saying that he (Lewontin) has demanded an unrealistic (mathematical) level of proof.
Lewontin makes no comment about the dysgenic trends discussed in his paper.
 The evidence Jensen uses for dysgenic trends is indirect. Geneticists who used to make these arguments were proven wrong by Carl Bajema, who showed that this argument was based on an egregious statistical error: they forgot to count women who had no children.
In the final analysis, the main point for Lewontin remains: even if the difference between black and white were entirely genetic, what program for social action would flow from that fact?
We are forced to examine all possible reasons for inequality. Society will benefit most if scientists and educators treat these problems in the spirit of scientific inquiry rather than as a battlefield upon which one of another preordained ideology may seemingly triumph.
 Nonsense. What we are morally obliged to do is to eliminate blackness as a cause of unequal treatment – and for that program we have no need of genetics. The decision about what role each child is to play eventually in society and what rewards he will receive is social.



5. Connections and resources, e.g., annotations to additional readings
American Psychological Association (2001). "New model of IQ development accounts for ways that even small environmental changes can have a big impact, while still crediting the influence of genes." www.apa.org/releases/iqmodel.html (Apr. 15).

Taylor, P. J. (2014) Nature-Nurture? No: Moving beyond the gaps in the sciences of variation and heredity" Arlington, MA: The Pumping Station—from page 41:

Woodhead, M. (1988). "When psychology informs public policy." American Psychologist 43(6): 443-454.

5b. Add to this blog post to provide suggestions and resources for revising the chapter.
5c. Adaptation of themes from the chapter to students' own projects of engaging your audience in learning or critical thinking about biology in its social context: Suggestions for how to do that:
  1. Identify writing where people speak about differences between groups as if everyone in each group can be treated as if they were at the group average. (See examples above from Taylor 2014.)
  2. Brainstorm with instructor