Posted by Meatwood_Flack on May 1, 2013, at 15:02:00
In reply to Re: Marijuana Not Medicine Addiction Docs Say » Meatwood_Flack, posted by SLS on May 1, 2013, at 7:57:41
Lawler, J.M. (1978). "Reliability and validity of IQ tests" (pp. 53-68). In I.Q. Heritability, and Racism. New York: International Publishers.
"Reliability" and "validity" of IQ tests
Test manuals describe the degree of "reliability" and the "validity" of the tests, which may suggest that IQ tests are "reliable" measures of intellectual capacity and have been validated by some criteria independent of the a priori conditions of testing.
The less important for our purposes is the question of "reliability," although this may be quite important from a technical point of view. More important is the question of "validity," a concept that seems to be more related to the notion that the tests have somehow been "proven" to measure "intelligence."
In addition to the "reliability" of tests due to minimizing "sampling error," described above, there are different ways of estimating the ranges of error which may be attributed to a test as a result either of short-term fluctuations in the psychological state of the individual taking the test or of chance familiarity or ignorance of material involved in specific test items. The second main aspect of reliability has to do with the manner in which the test is administered.
Reliability
Without going into detail regarding methods of determining reliability, it is enough to note that this concept refers to the elimination of short-term variations in the performance of the child or the [p.54] tester. In introducing the subject of "mental tests," one psychologist writes, "Mental testing is simply a more refined and scientific method of doing something which we all do every day of our lives, that is assessing one another's abilities and character traits... For example, we might watch a small boy playing with bricks, and from his skillfulness, or the elaborateness of the resulting construction, jump to the conclusion that he is quite a bright child. Obviously the basis for this and other such judgments is very haphazard and unreliable."
It is one thing for an individual to guess the relative intelligence of a child, it is another matter for many individuals to make such a guess. IQ tests basically reflect the average opinion or "representative" judgment of teachers regarding the relative performance of a child compared to children of the same age for the entire nation. It is in this sense that the tests are more reliable than the snap judgment of a single person. But we would not agree that the tests are thereby more scientific. There may be continuity between the intuitive judgment of the individual and the collective judgment of society, but it is still a matter of a subjective judgment operating within the narrow constraints of a definition of intelligence which is itself far from complete and scientific.
Statistical methodology may make an opinion poll a far more "reliable" indicator of the general opinion of a nation than a brief conversation with one's next door neighbor. This does not make the opinions expressed through the opinion poll truer in themselves, although they may be more representative than the opinions of one's neighbor. This point is considerably obscured, however, in the case of the opinion polling regarding intelligence which is at the basis of IQ. The methodology used may make such results more representative and "reliable" than the snap judgment of an individual as a way of "measuring" the intelligence of a particular child. But when some psychologists boast that IQ tests are more reliable predictors of an individual's performance than are judgments based on particular interviews, it should not be thought that this fact, if it is generally true, proves that a scientific way of measuring general intellectual capacity has been discovered. IQ tests are only a "refined" form of the relative estimate of intelligence made by the common sense judgment of teachers and psychologists. [p.55]
Validity
The term "validity" in psychometric theory implies more than the simple reliability of test scores. Not only are such scores freed from chance variations in the mood of the child or the tester, but they are said to be validated by some criterion independent of the test construction as actually measuring "intelligence." The independence of the criterion for validating a measurement is crucial to proving that one has not simply forced an object into the arbitrary confines of an a priori conception of intelligence and system of measuring it. It is often possible to read what we want to into a particular event or character. Often a person's appraisal of another person tells us more of the character of the first person than of the second. If the person who makes the judgment is consistent we can attribute a great deal of "reliability" to such judgments. We can rely on this person to make the same biased judgments in every case. But we recognize that this reliability does not make the judgments valid, i.e., true. The judgments are validated by some criterion that is independent of the mere fact of the judgment or of its short-term consistency. Can IQ tests be validated by some criterion that is independent of the fact that a concept and method of measuring IQ is applied consistently and "standardized?"
Four ways of determining validity are generally recognized: content, predicative, concurrent and construct validity. A brief description and commentary on these four ways of validating tests will further clarify the nature of IQ tests and the technical theory of test construction. We will see that none of these criteria can truly be said to be independent of the way tests are constructed and to prove that the tests in fact measure "general cognitive ability."
Content validity
"Content validity" refers to the problem of whether a test "covers a representative sample of the behavior domain to be measured." Thus a test that is supposed to measure musical ability should be composed of more than tests of an individual's knowledge of major composers and their works. To measure an individual's musical ability it is necessary to understand various relevant aspects [p.56] of this ability, and provide tests which demonstrate competency in each of these aspects.
The concept of content validity implies that one understands the ability in question and can analyze its component parts in such a way as to construct a test that accurately "samples" the various features of that ability. Real demonstration of the ability, in the degree predicted, would seem to be the ultimate criterion of the validity of the test. In the case of the IQ test, however, there is no generally agreed definition of the area of "general cognitive ability," if that is what the IQ is supposed to measure. Thus the content of the tests was not derived by an analysis of the "behavior-domain," general cognitive ability, but simply by selecting items that discriminate children on the basis of their school performance but which do not appear to be directly learned through school. The "content validity" seems to consist in the fact that the items look like items that measure "general intelligence" as defined by intuition.
But this is more a matter of giving plausibility to the test than of analyzing an objective domain into its component parts. Anne Anastasi notes that IQ tests have no clear validity on this score: "Aptitude and personality tests bear less intrinsic resemblance to the behavior domain they are trying to sample than do achievement tests. Consequently the content of aptitude and personality tests can do little more than reveal the hypotheses that led the test constructor to choose a certain type of content for measuring a specified trait. Such hypotheses need to be empirically confirmed to establish the validity of the test." In other words, we need to look elsewhere to confirm the hypothesis which led to the selection of the types of items that appear in IQ tests.
Tests which effectively measure the existence of certain abilities, and are proven to be "valid" by actual performance involving the ability are still not scientific in the full sense of the term. David McClelland, who argues that IQ tests are simply measures of school standing and not of innate capacity, calls for tests based on real "competence." He hopes thereby to eliminate the invidiousness of so-called intelligence tests which assume general intellectual incapacity. Testing for "competence" rather than "intelligence" may eliminate many of the abuses connected with IQ tests. However, as long as the causes of the development of the abilities in question are not explained, tests remain a matter of measuring already formed [p.57] abilities. Structural biases in the development of "competencies" would still be reflected in the use of such tests. Moreover, hereditarian explanations of the more exactly defined "competence" are not eliminated by such reforms. McClelland attacks the hereditarian theory of IQ in a manner similar to Anastasi, i.e., within the framework of "sorting and selecting," but not of explaining development so as to be able to master the processes involved in various social and intellectual skills.
Predictive validity
The second type of validity ascribed to IQ tests is "predictive validity." Of the various criteria of validity this one seems to be the most significant. As a general rule IQ scores are relatively stable. Especially after the age of six or so, an individual's IQ score remains relatively constant. Consequently, IQ scores are good predictors of relative school performance. However it is not surprising that IQ tests predict school performance to a significant degree. This is only a reflection of the fact that once a rank order is established in any school it tends to be relatively stable. Children who are "at the top" in their early years tend to stay there, while children at the bottom remain there as well. This is true not only within schools, but, perhaps even more importantly, between schools. Good schools generally remain good schools, and poor schools stay poor. If IQ tests are constructed to reflect the rank order of performances at each age level in relation to a national average, there is no mystery to the relative stability of IQs and to the predictive power of IQ tests regarding school achievement. This is nothing other than a reflection of a well-known fact of life in the schools -as well as between the schools. This stability of the rank order of school performances is therefore not explained by IQ, since the stability of IQ is basically a reflection of this stability of relative performance.
It is not necessary to hypothesize differences in fixed mental capacity to explain the degree of fixity of rank order in school success. Differentiation between children is not simply a spontaneous occurrence in schools where everyone is "treated equally." Not only are there major differences between schools, but within schools themselves there is a system of "tracking" or "streaming" children which tends to guarantee a stability of rank order. [p.58]
At the most general level, differentiation between individuals in school is guaranteed by a system of instruction which places primary emphasis on competition between individuals -rather than, for example, on cooperative efforts and joint projects of children, and/or competition between groups of children.
Closely bound up with the general feature of individual competition in the classrooms is the tracking of children in levels of achievement. The tracking system tends to rigidify rank order. Although tracking is often justified on the ground that children have different basic capacities, this justification and explanation is more in the order of a self-fulfilling prophecy -as is the predictability of IQ tests. Even if IQ scores themselves play no role in determining expectations -both in the teacher and quite soon in the pupil as well- relative judgments of individual standing would nevertheless form and solidify the rank order of children in schools. Common practices, such as placing the best readers in the first row, combined with disciplinary measures, rewards and punishments, etc., have substantial influence over an individual's eventual performance and "self-image." Deeper studies of the psychological and pedagogical effects of individual competition and rank ordering according to general levels of performance would shed much light on the question of explaining relative stability of rank ordering.
Of primary importance in an examination of the dynamics of learning is the presence of racism, creating an oppressive climate for minority children and acting to undermine educational motivation for children, many of whom see little encouragement or opportunity to develop useful academic skills. The existence of a basically segregated school system in the United States constitutes a national tracking system in which race, and not IQ, is clearly of primary importance.
Segregated schooling, class differences between schools, the tracking system, and competitive evaluation of individual performance constitute a complex, practical system of education. This system is often justified on the grounds that there are innate, individual, class and racial differences in capacity. Since, it is argued, the schools provide "equal opportunity," the differences in performance must be due to differences in either "basic ability" or "effort" in individuals themselves. In fact the schools offer no such equal opportunity, but are themselves based on a concept of differential [p.59] education. The fact that primary schools are more homogeneous in form than secondary schools gives the impression that for some eight years children are given an equal chance to "show what they are made of," and only with secondary school are they finally tracked along vocational or academic lines. A deeper analysis of practices on the primary school level shows that this "fairness" is superficial. All levels of education involve unequal relations between teachers and students, and between the children themselves as they internalize their own relative degrees of academic inferiority and superiority.
Thus "predictive validity" essentially raises more questions than it answers. It does not refute our claim that "intelligence" and the IQ methods for measuring it rest on certain presuppositions which are not independently verified. Since the process of verification itself remains within the sphere of these propositions, there is no proof by correlations with later school performance that IQ tests measure general intellectual capacity.
Concurrent validity
The third form of "validity" is concurrent validity. This concept has to do with the validation of the test, either by the actual performance which the test is meant to measure, or by correlation of the scores with the scores of other tests meant to measure the same thing.
As to correlation with other test scores this involves an obvious begging of the question. For the question remains as to the basis of the validity of these other tests. The widespread practice of "validating" other tests, particularly group tests, by the correlation of scores with the Stanford-Binet is only a testimony to the importance of the latter. It hardly shows the validity of the other tests as measures of "general intellectual capacity."
Under the heading of concurrent validity Anastasi lists using academic achievement and years of schooling, teachers' ratings or the ratings of psychologists and trained observers, as well as the method of "contrasted groups." Under "ratings" Anastasi states that "personal judgments" can be the "very core of the criterion measures. Under these circumstances the ratings themselves define the criterion." The method of contrasted groups is simply the [p.60] method of selecting "bright" and "dull" groups of children, and measuring individual items and the test as a whole as to their "validity" in distinguishing these opposed groups. Anastasi stresses the use of groups which have been sorted out by "the cumulative and uncontrolled selective influences of everyday life.... The criterion is ultimately based upon survival within a particular group versus elimination therefrom." An example of this is the contrast of "institutionalized mental defectives" and normal school children of the same age. (Presumably, children in the first group have not been institutionalized as a result of their performance on IQ tests.)
The use of "survival of the fittest" theory to give validity to IQ test harkens back to Social Darwinism. In developing tests of adult intelligence Terman used "contrasting groups" of 30 business men, 150 "'migrating' unemployed men," 150 adolescent delinquents and 50 high school students. The assumption in this selection seems to be that "life itself" has selected these individuals on the basis of their innate intelligence. The tests should thus reflect the superiority of the businessmen over the unemployed, and the high school students over the "adolescent delinquents." Terman does not go into any detail regarding his method of selecting items, noting only that "Adults whose intelligence is known from other sources to be superior are found to test well up toward the 'superior adult level,' and this holds whether the subjects in question are well educated or practically unschooled. The almost entirely unschooled business men, in fact, tested fully as well as high-school juniors and seniors." The fact that the "unschooled" businessmen scored as well as the high school student is obviously understood to mean that these tests reflect "native" intelligence.
Although no information is given here on how items are selected, it seems clear that this "finding" is actually a premise for the construction of tests meant to show the native intelligence of economic survivors to be superior to life's dropouts. Items could presumably be found which the delinquent adolescents and the migrating unemployed could answer, and which the unschooled businessmen and (their?) schooled adolescents could not.
Thus "concurrent validity" for the Stanford-Binet consists in the agreement of the test with judgments of intelligence made by school teachers, with survival in school versus dropping out, and [p.61] success in life versus unemployment. But these criteria are a basis of test construction, not a confirmation of the validity of the test constructed on some other basis. The justification of this procedure is the a priori notion that success in school and life is determined by innate intelligence and therefore tests which distinguish the successful (in school and in life) from the successful will have measured innate intelligence.
Construct validity
The fact that these three methods of validating IQ as a measure of general intelligence fail to do so leads to a fourth, murkier, and more clearly subjective theory of validation -so-called construct validity. According to Anastasi, "The construct validity of a test is the extent to which the test may be said to measure a 'theoretical' construct or trait." Such a "construct" as "intelligence" involves a "broader, more enduring and more abstract kind of behavioral description than the previously discussed types of validity" and so requires "the gradual accumulation of information from a variety of sources. Any data throwing light on the nature of the trait under consideration and the conditions affecting its development and manifestations are grist for this validity mill. As illustrations of the specific techniques utilized may be mentioned age differentiation, correlations with other tests, factor analysis, internal consistency, and effect of experimental variables on test scores."
Essentially, the above description of construct validity underlines a kind of scientific limbo in which the problem of validation places IQ tests. Anything that seems to give some support to the vague concept of intelligence is grist for this last, catchall validity mill. Jensen's method of proving the innateness of IQ consists precisely in accumulating "various lines of evidence, not one of which is definitive alone, but which viewed all together, make it a not unreasonable hypothesis that genetic factors are strongly implicated in the average Negro-white intelligence difference."
There is no need to elaborate on the matter of age differentiation which is clearly a condition of the construction of the test, and not an independently discovered correlation. Anastasi admits that various correlations with other tests are similar to techniques discussed under concurrent and content validity. [p.62]
How does "factor analysis" demonstrate the "validity" of the tests"? According to Anastasi, factor analysis first of all simplifies the test by breaking it down into a small number of categories from the originally large number of particular items. Each "factor" -such as vocabulary, verbal comprehension, analogies, etc.- is then correlated with the overall test in order to determine the "factorial validity" of that particular factor. Thus, for example, if children, who have an average IQ happen also to have average scores on the vocabulary tests, but above average (or below average) scores on analogies, the vocabulary subtest is said to have a higher factorial validity than the analogies subtest. If the score on the subtest is the same as the score on the entire test, the subtest has a factorial validity of 1.00 -perfect correlation. If the factorial validity of a subtest is very low -showing little correlation with the test as a whole- it is then doubtful whether the subtest belongs in the test at all. Thus factorial validity is a method of validating particular groupings of items by comparing them with the test as a whole. Instead of a validity factor outside of and independent of the test, the test itself is here regarded as the criterion of the validity of the factors that compose it.
Jensen's arguments of validity: "g" and Headstart
Jensen regards Charles Spearman's correlations of the subtests to be an especially cogent argument that the tests really measure a basic "source of variation" of performance, some "general intelligence," or simply "g." Thus, according to Jensen,
Spearman noted that if the tests called for the operation of "higher mental processes," as opposed to sheer sensory acuity, reflex behavior, or the execution of established habits, they showed positive intercorrelations, although the tests bore no superficial resemblance to one another... For example, a vocabulary test shows correlations in the range of .50 to .60 with a test that consists of copying sets of designs with colored blocks; and a test of general information correlates about .50 with a test that involves winding through a printed maze with a pencil.... To account for the intercorrelations of "mental" tests, he hypothesized the existence of a single factor common to all tests involving complex mental processes... Spearman called the common factor 'general intelligence' or simply g. [p.63]
Thus, for Jensen, factor analysis is a method of validating IQ tests because it breaks the test down into a number of different factors, and then demonstrates that all of these factors "intercorrelate" to a significant degree. Does this high intercorrelation prove that IQ tests measure "general intelligence"? Here again, a little reflection on the method of test construction shows that there is no independent criterion of validation of the IQ score. Intercorrelation of the items is in fact a condition for the construction of the test. Items which the "bright" children do not pass are not considered to be good items for distinguishing "bright" from "dull" children at a particular age. Such items are either eliminated or placed somewhere else on the age scale where they "work." Thus all the items of the test must by definition "intercorrelate" as much as is possible given the empirical, trial and error manner of test construction.
Jensen does not explain this matter of the a priori character of item intercorrelations. However, he makes a "statistical qualification" for the benefit of those who might understand that "g" is an artifact of test construction: "We should not reify g as an entity, of course, since it is only a hypothetical construct intended to explain covariation among tests. It is a hypothetical source of variance (individual differences) in test scores." In fact "g" cannot explain anything. The fact of intercorrelations is itself explained by the method of test construction. It is Jensen who hypothesizes that the "hypothetical construct," etc: "Despite numerous theoretical attacks on Spearman's basic notion of a general factor, g has stood like a rock of Gibraltar in psychometrics, defying any attempt to construct a test of complex problem solving which excludes it."
The a priori artificial character of "g" is brought out by Anastasi, when she examines "internal consistency" as a method of validating tests. Thus, in "one method of constructing a test two groups of children are identified, one judged to be "bright," the other "dull":
Items that fail to show a significantly greater proportion of "passes" in the upper than in the lower criterion group are considered invalid, are either eliminated or revised. Only those items yielding significant item-test correlations would be retained. [p.64]
Thus correlations between scores on the test as a whole and either individual items or groups of items is a necessary condition of test construction. The fact that factor analysis "finds" this general consistency in the test, despite the relative diversity of items which make up the test, is not at all surprising and does not demonstrate that the tests measure "general intelligence." At best it is a reflection of the fact that children who do better than others in school do so in a number of relatively different areas of mental activity.
As to "effect of experimental variables on test scores," this method of construct validation refers to experiments designed to confirm the validity of the scores by retesting under different conditions.
Experimental testing of the construct of IQ as a fixed intellectual capacity would consist in "interventions" in the normal school system to see whether such interventions would change IQ scores. Thus, if significant variation of the school system does not change the rank ordering of the children, then the concept that IQ measures a basically fixed capacity would then appear to be at least partially confirmed. When Jensen refers to Operation Headstart as a "failure," he is arguing that just such experimental variation of the school system had in fact occurred, and had failed to "boost" the IQs of the children involved. However, as Jensen himself admits, IQ scores were in fact affected positively by Headstart. Jensen argues that this boosting was only temporary and disappeared after two years. However, the fact that after two years of returning to the "normal" situation children's IQs or rank order returned to "normal" does not mean that Headstart failed. Only if the intervention was supposed to affect a "general intelligence" or "g" that remains at the basis of all learning could it be said that Headstart failed to affect this "hypothetical construct." If IQ tests simply involve the testing of skills and ideas somehow involved in school performance at a particular age, an intervention program which increased performance in those skills would not directly affect the different abilities which represent normal performance at a later age.
Headstart only failed to confirm Jensen's own hypothesis that intelligence is an underlying "source of variance" of school performance, which if affected by an "environmental" effort should remain durably affected. Our argument, on the contrary, is that this definition of intelligence is theoretically and empirically unjustifiable. [p.65] The fact that a temporary intervention should have only temporary success is hardly proof that it is a failure. The limited character of both the effort and the results naturally suggests the need for a more expanded program. It is in the face of this perspective that Jensen argues that the experiment failed to boost "intelligence" -an admittedly "hypothetical construct"- while "only" improving more superficial "achievement."
In conclusion, none of the standard arguments for validating IQ proves anything other than that these scores reflect rank order in school performance and a relative stability of this rank order. There is no validation of the "constructs" "general intelligence," "cognitive capacity" or, most importantly, "innate, general cognitive capacity." All these "constructs" remain simply that -constructions of the mind, a priori ideas that have guided the development of a test that seems to make these ideas plausible. In the light of the above discussion it is unclear what the concept of "validity" really means for IQ tests. At best it seems to mean that the tests have in fact been constructed according to the rules laid down for their construction. Aside from this there appears to be no "validation" of these rules, and the concepts or hypotheses standing behind them, by a practical result which is not essentially presupposed by the test in the first place.
Behaviorist interpretation
Although Anastasi, for example, seems to support the methodology of validation as demonstrating that IQ tests are valid measures of "something," her own arguments rule against the notion that any of the major concepts of intelligence testing are valid. Thus, in a discussion of construct validation, Anastasi argues that no hard and fast line can be drawn between construct validation and the other forms of validation. In fact the latter might all be included under the broadly defined notion of construct validation.
Anastasi concludes with a warning against speculation on the meaning of various kinds of tests:
It would also seem desirable to retain the concept of the criterion in construct validation, not as a specific practical measure to be predicted, but more generally to refer to independently gathered external data. The need to base all validation on data, rather than on armchair speculation, would thus be reemphasized, as would the need for data external to the test scores themselves. Internal analysis of the test, through item-test correlations, factorial analyses of test items, etc., is never an adequate substitute for external validation.
So far we have failed to find any external validation of IQ scores as measures of anything other than rank order of school performance, which is what IQ tests are based on in the first place.
It is in the spirit of opposing armchair speculation that Anastasi opposes the interpretation of IQ tests as measures of general cognitive capacity. As to the content of the tests, she opposes the concept that the items used are examples of unlearned abilities, qualitatively different from achievement tests:
The increasing efforts to prepare achievement tests that would measure the attainment of broad educational goals, as contrasted to the recall of minutiae, also made the content of achievement tests resemble more closely that of intelligence tests. Today the difference between these types of test is chiefly one of degree of specificity of content and extent to which the test presupposes a designated course of prior instruction.
As school instruction gives even greater attention to general patterns of cognitive development, even this distinction is beginning to disappear. "Spontaneous" operations of thought have been recognized to be products of definite conditions of development and subject to educational measures. This is all the more true as these operations of thought become more complex and reach the level of modern scientific theories and methods. An arbitrary line between basic unlearned and general intellectual abilities, and derived, acquired and specific abilities cannot be drawn -at least not in the way supposed by IQ ideology.
As to the meaning of the predictive validity of IQ scores, Anastasi warns against explanations of this fact from basic differences in "capacity":
Only in the sense that a present behavior sample can be used as an indication of other, future behavior can we speak of a test measuring "capacity." No psychological test can do more than measure behavior. Whether such behavior can serve as an effective index of other behavior can be determined only by empirical tryout. [p.67]
Thus, from this behaviorist viewpoint, speculation on the meaning of IQ scores appears unwarranted. A particular "behavior sample" simply correlates with other behavior. No inference can be drawn from observed behavior to unobserved "capacity." No assertions can be made about this behavior sample other than that it seems to be somewhat more general and less explicitly the object of school instruction than other "behavior samples" included in achievement tests. The behaviorist standpoint taken by Anastasi represents, on the one hand, a response to the mounting criticism against the pretensions of IQ tests. On the other hand, it represents an attempt to preserve them within an [ontologically-PB] "agnostic" and [operationalized-PB] positivist philosophical framework. Aware of the abuses of IQ tests, Anastasi attempts to remove the philosophical, "metaphysical" or speculative interpretations of IQ tests while preserving them on utilitarian grounds. After all they do correlate with other behavior, whatever the reason. Thus, on the use of IQ tests in the early days, she writes that:
The application of such group intelligence tests far outran their technical improvement. That the tests were still crude instruments was often forgotten in the rush of gathering scores and drawing practical conclusions therefrom. When the tests failed to meet unwarranted expectations, skepticism and hostility toward all testing often resulted. Thus the testing boom of the twenties, based upon the indiscriminate use of tests, may have done as much to retard as to advance the progress of psychological testing.
Between the materialistic and developmentalist explanation of intelligence and the idealistic interpretation of intelligence as a fixed inner capacity stand the intermediate philosophical position of agnosticism and empiricism, represented here by behaviorist defense of IQ tests without pretending to "speculate" on their meaning. Directed against the hereditarian interpretation of IQ, which is being defended by Jensen today, it is also directed against critics of IQ tests who argue that they are inherently loaded with unwarranted conceptions. On the other hand, the criticism of the fallacies of IQ ideology serves to "demythologize" the IQ test, at least in the minds of some specialists. But the agnostic position which consists in refraining from explanations, while simply noting associations of data or statistical correlations, leaves the door open to Jensen and others who express impatience with such "fastidiousness." The [p.68] attempt to provide an explanation or to propose hypotheses is inherent in the process of scientific development. Jensen takes advantage of this fact in challenging both the agnostic refusal to put forward an explanation and environmentalist explanations generally. Moreover, since many psychologist use behaviorist [(i.e., operationist or interactionist -PB)] concepts and methods, they are poorly prepared philosophically to defend the environmentalism they may in fact espouse.
The point of view which we are defending here is not one which seeks to "retard... the advance of psychological testing." By examining the presuppositions of IQ testing we are attempting to show that this type of psychological test has a very limited value which is basically far outweighed not only because of "unwarranted expectations," but because of features that are inherent in its essential make up.
*See also Lawler's (1978) Introduction; and piercing critique of Jensen's "heritability estimates." In the latter, Lawler also deconstructs the implied assumptions of the more typical "'interactionist' position" on testing and human intellect. He hints successively that the "point of view" mentioned in the last paragraph (immediately above) is one which necessitates that a transformative sociohistorical approach to the development of human intellect must be adopted. In doing so, Lawler supplies us with a rare disciplinary exemplar of someone who attempts to describe or explain human intellectual development from outside the self-serving "framework of 'sorting and selecting'" (p. 57).
Anastasi, A. (1961). Psychological Testing, The Macmillan Company, New York.
Jensen, A. (1969). "How much can we boost IQ and scholastic achievement?" Harvard Educ. Rev., 39, 1-123.
McClelland, D. (1974). "Testing for Competence Rather Than for 'Intelligence,'" in Gartner, Greer and Riessman, eds., The New Assault on Equality, Perennial Library, New York.
Paul F. Ballantyne, Ph.D. Posted:[June, 2007]
pballan@comnet.ca
poster:Meatwood_Flack
thread:1042969
URL: http://www.dr-bob.org/babble/20130501/msgs/1043036.html