This article was downloaded by: [Universitara M Emineescu Iasi] On: 17 November 2011, At: 05:17 Publisher: Psychology

Press Informa Ltd Registered in England and Wales Registered Number: 1072954 Registered office: Mortimer House, 37-41 Mortimer Street, London W1T 3JH, UK

Journal of Clinical and Experimental Neuropsychology

Publication details, including instructions for authors and subscription information: http://www.tandfonline.com/loi/ncen20

Age-related IQ decline is reduced markedly after adjustment for the Flynn effect
Mercedes D. Dickinson & Merrill Hiscock
a b a a b

Department of Psychology, University of Houston, Houston, TX, USA

Center for Neuro-Engineering and Cognitive Science, University of Houston, Houston, TX, USA Available online: 26 Mar 2010

To cite this article: Mercedes D. Dickinson & Merrill Hiscock (2010): Age-related IQ decline is reduced markedly after adjustment for the Flynn effect, Journal of Clinical and Experimental Neuropsychology, 32:8, 865-870 To link to this article: http://dx.doi.org/10.1080/13803391003596413

PLEASE SCROLL DOWN FOR ARTICLE Full terms and conditions of use: http://www.tandfonline.com/page/terms-and-conditions This article may be used for research, teaching, and private study purposes. Any substantial or systematic reproduction, redistribution, reselling, loan, sub-licensing, systematic supply, or distribution in any form to anyone is expressly forbidden. The publisher does not give any warranty express or implied or make any representation that the contents will be complete or accurate or up to date. The accuracy of any instructions, formulae, and drug doses should be independently verified with primary sources. The publisher shall not be liable for any loss, actions, claims, proceedings, demand, or costs or damages whatsoever or howsoever caused arising directly or indirectly in connection with or arising out of the use of this material.

JOURNAL OF CLINICAL AND EXPERIMENTAL NEUROPSYCHOLOGY 2010, 32 (8), 865870

NCEN

Age-related IQ decline is reduced markedly after adjustment for the Flynn effect
Mercedes D. Dickinson1 and Merrill Hiscock1,2
Department of Psychology, University of Houston, Houston, TX, USA Center for Neuro-Engineering and Cognitive Science, University of Houston, Houston, TX, USA
2 1

Age-related IQ decline

Downloaded by [Universitara M Emineescu Iasi] at 05:17 17 November 2011

Twenty-year-olds outperform 70-year-olds by as much as 2.3 standard deviations (35 IQ points) on subtests of the Wechsler Adult Intelligence Scale (WAIS). We show that most of the difference can be attributed to an intergenerational rise in IQ known as the Flynn effect. Normative data from different versions of the WAIS enabled us to estimate the degree to which the Flynn effect, rather than age-related decline, contributes to differences between 20- and 70-year-olds. The Flynn effect accounted for 3867% of the apparent age-related decline on 6 of the 11 subtests. On the other 5 subtests, all of which are categorized as verbal, the Flynn effect was larger than the age-group difference. For these verbal subtests, the Flynn effect masked a modest increase in ability as individuals grow older. Overall, the Flynn effect accounted for at least 85% of the disparity between 20- and 70-year-olds. Keywords: Aging; Flynn effect; Cohort effect; Gerontology; Intelligence; IQ testing.

Mean intelligence quotient (IQ) remains constant throughout adulthood only because IQ test norms are age adjusted to ensure that the average for each age group is 100 (Wechsler, 1955). Normative raw scores differ between younger and older adults, with the differences favoring the younger groups. On IQ tests that comprise multiple components, such as the Wechsler Adult Intelligence Scale (WAIS), the magnitude of age differences varies substantially from one subtest to another. Nonverbal subtests typically show larger age differences than do verbal subtests (Wechsler, 1955, 1981, 1997). These age differences suggest that cognitive abilities decline beyond the age of 50 years or so, and that some abilities decline faster than others. An alternative explanation is that the apparent decline is an artifact of the intergenerational rise in IQ known as the Flynn effect (Flynn, 1984, 1987, 1998a, 1998b, 1999, 2006, 2007; Hiscock, 2007; Neisser, 1998). In other words, cognitive abilities appear to be diminished in older people because subsequent generations are becoming increasingly more intelligent. We tested this alternative explanation by

comparing the magnitude of the age-related decline in IQ test raw scores with the magnitude of the Flynn effect on the same raw scores. Age-group norms for Wechsler component tests are cross-sectional, which means that change over time is not measured directly but must be inferred from differences between groups of different ages. Even when age groups are carefully matched for demographic characteristics, the cross-sectional design is vulnerable to cohort effects, which are differences other than age per se between younger and older samplesfor example, differences in child-rearing practices, health-care services, economic and political conditions, and educational practices (Schaie, 1986, 1994). Specific cohort effects are difficult to isolate and measure, but their aggregate impact on IQ has been quantified during the past 25 years in studies of the Flynn effect (e.g., Flynn, 1984, 1987, 1999, 2006; Neisser, 1998). The Flynn effect may be understood as a cohort (or birth date) effect, as it favors people born at a later time even if age at testing is held constant. This phenomenon

The authors thank James R. Flynn for providing prepublication drafts of articles and other information that facilitated the writing of this paper. We appreciate the suggestions received from Arturo Hernandez and other members of the developmental cognitive neuroscience faculty of the Department of Psychology, University of Houston. Address correspondence to Merrill Hiscock, Department of Psychology, 126 Heyne Building, University of Houston, Houston, TX 772045022, USA (E-mail: mhiscock@uh.edu).

2010 Psychology Press, an imprint of the Taylor & Francis Group, an Informa business http://www.psypress.com/jcen DOI: 10.1080/13803391003596413

866

DICKINSON AND HISCOCK

Figure 1. Schematic diagram of basic method.

has been documented in various parts of the world and has been occurring since the end of the nineteenth century (Flynn, 1999, 2007). The magnitude of the largely linear IQ increase is 3 to 6 IQ points per decade, depending on the test used to measure IQ (Flynn, 1984, 1987). The factor or factors responsible for this pervasive rising curve are unknown but are assumed to be environmental because no known genetic mechanism can account for the large IQ differences that occur between successive generations of the same population (Dickens & Flynn, 2001; Flynn, 2007). We sought to quantify the true effect of aging on IQ by adjusting cross-sectional age-group norms for the Flynn effect. Each of the 11 Wechsler subtests was analyzed separately, and adjustments were calculated separately for the 1981 Wechsler Adult Intelligence Scale Revised (WAISR) and the 1997 Wechsler Adult Intelligence ScaleThird Edition (WAISIII). Experiment 1 applies Flynn-effect adjustments to the 1981 Wechsler test, and Experiment 2 similarly adjusts the 1997 revision of the WAIS. In Experiment 3, we combine data from Experiments 1 and 2 to construct an optimized model of age-group differences before and after Flynn-effect adjustment. The general strategy is depicted in Figure 1. First we determined the difference in scaled-score units between 20- and 70-year-olds for each of the 11 subtests of the Wechsler IQ test. Then we adjusted those differences for the Flynn effect as calculated for each subtest and extrapolated to a 50-year interval. All calculations were based on data contained in the administration or technical manuals for the respective version of the Wechsler test. EXPERIMENT 1 Method Three steps were required to adjust the WAISR agegroup norms for the Flynn effect. First, average performance for 70-year-olds on each of the 11 WAISR subtests was expressed as a scaled score based on the distribution for 20-year-olds. Scaled scores have a mean of 10 and a standard deviation of 3. For each subtest, the raw score that would yield an average scaled score (i.e., a

Downloaded by [Universitara M Emineescu Iasi] at 05:17 17 November 2011

scaled score of 10) for 20-year-olds in the standardization sample was determined. This average raw score then was applied to the norms for 70-year-olds to determine the scaled score that a 70-year-old would receive if he or she scored at the mean for 20-year-olds. For example, if a raw score of 18 yielded a scaled score of 10 for 20-yearolds, the same raw score might yield a scaled score of 13 for 70-year-olds. Thus, performance that is average for 20-year-olds would be one standard deviation above average for 70-year-olds. This would indicate that the mean scaled score for 70-year-olds on that subtest is 3 scaled-score units, or one standard deviation, below the mean for 20-year-olds. The mean difference between the two age groups, which we refer to as the age-group difference (AGD), is assumed to reflect the true effect of aging plus the Flynn effect. The AGD is expressed in terms of scaled-score points per 50 years. Step 2 entailed calculating the magnitude of the Flynn effect on each of the subtests. Calculations were based on data from a validity sample of 72 individuals who completed both the WAIS and the WAISR in 1978 (Wechsler, 1981). To the degree that the population average had risen over time, the newer test (WAISR) would yield lower scaled scores than the previous test (WAIS). The interval between WAIS and WAISR standardization dates was 24.5 years (Flynn, 1999). The magnitude of the mean difference in scaled scores for each subtest, when extrapolated to a 50-year interval, indexes the strength of the Flynn effect for that subtest. We call this score the Flynn effect difference (FED). In Step 3, a true aging effect (TAE) was calculated as the algebraic difference between the FED and AGD. The TAE represents the AGD after adjustment for the Flynn effect. The units are scaled-score points per 50 years. Results and discussion Table 1 shows the WAISR results in scaled-score points per 50 years. Averaged across all 11 subtests, the 50-year AGD is 3.3 scaled-score units (16.5 IQ points), and the comparable FED is 2.8 scaled-score units (14.0 IQ points). Thus, the Flynn effect is sufficiently large as to account for 85% of the difference between age groups. By subtraction, the TAE is 0.5 scaled-score units, which represents an actual decrease of only 2.5 IQ points over a 50-year interval. Separate calculations of TAE for the six verbal and five performance subtests indicate that verbal scores increase by 7.0 IQ points, whereas performance scores decrease by 13.5 IQ points. It was anticipated that the AGD would be larger than the FED in all instances, which would indicate a residual age-related intellectual decline after adjustment for the Flynn effect. The results for the performance subtests and one verbal test (Digit Span) conform to this pattern. However, in the case of the other five verbal subtests, the Flynn effect is larger than the age-group difference, which indicates that the Flynn effect is masking an actual increase in raw scores between the ages of 20 and 70 years.

AGE-RELATED IQ DECLINE TABLE 1 Calculation of true aging effect for WAISR subtests FED from WAIS to WAISRa 2.2 1.2 3.7 2.0 3.7 4.5 3.7 1.6 2.0 2.7 3.7 2.9 2.7 2.8 TABLE 2 Calculation of true aging effect for WAISIII subtests FED from WAISR to WAISIIIa

867

Subtest Verbal subtests Information Digit Span Vocabulary Arithmetic Comprehension Similarities Performance subtests Picture Completion Picture Arrangement Block Design Object Assembly Digit Symbol Verbal mean Performance mean Overall mean

AGDa 1.0 2.0 1.0 1.0 1.0 3.0 5.0 5.0 5.0 5.0 7.0 1.5 5.4 3.3

TAE (AGD FED)a

Subtest Verbal subtests Information Digit Span Vocabulary Arithmetic Comprehension Similarities Performance subtests Picture Completion Picture Arrangement Block Design Object Assembly Digit Symbol Verbal mean Performance mean Overall mean

AGDa

TAE AGD FEDa

+1.2 0.8 +2.7 +1.0 +2.7 +1.5 1.3 3.4 3.0 2.3 3.3 +1.4 2.7 0.5

+1.0 2.0 +1.0 0 +1.0 1.0 3.0 5.0 4.0 4.0 6.0 0.0 4.4 2.0

0 0.3 1.8 +0.9 1.5 2.6 1.2 1.8 2.1 2.6 3.5 0.9 2.2 1.5

+1.0 1.7 +2.8 0.9 +2.5 +1.6 1.8 3.2 1.9 1.4 2.5 +0.9 2.2 0.5

Downloaded by [Universitara M Emineescu Iasi] at 05:17 17 November 2011

Note. AGD = age-group difference. FED = Flynn effect difference. TAE = true aging effect. WAISR = Wechsler Adult Intelligence ScaleRevised. WAIS = Wechsler Adult Intelligence Scale. aIn scaled score points per 50 years.

Note. AGD = age-group difference. FED = Flynn effect difference. TAE = true aging effect. WAISR = Wechsler Adult Intelligence ScaleRevised. WAISIII = Wechsler Adult Intelligence ScaleThird Edition. aIn scaled score points per 50 years.

EXPERIMENT 2 Method In Experiment 2 we applied the procedure used in Experiment 1 to data from the WAISIII (Wechsler, 1997). First we determined the difference in scaledscore units between 20- and 70-year-olds for each of the 11 subtests. This again was termed the age-group difference (AGD). Then we adjusted those differences for the Flynn effect as calculated for each subtest and extrapolated to a 50-year interval. The Flynn effect difference (FED) for each subtest in this instance was based on a validity sample of 192 people who completed both the WAISR and the WAISIII in 1995 (Tulsky, Zhu, & Ledbetter, 1997). Following Flynn (1999), we used 17.0 years as the interval between standardization dates for the respective versions of the WAIS. Finally, a true aging effect (TAE) was calculated as the difference between the FED and AGD. Results and conclusions Results for the 1997 WAISIII are shown in Table 2. The 50-year AGD is 2.0 scaled-score units (10.0 IQ points), and the comparable FED is 1.5 scaled-score units (7.5 IQ points). Thus, the Flynn effect accounts for 75% of the difference between age groups. The TAE, by subtraction, is 0.5 scaled-score units, which indicates a decrease of 2.5 IQ points between the ages of 20 and 70

years. TAE for verbal subtests indicates an increase of 4.5 IQ points, and TAE for performance subtests indicates a decrease of 11.0 IQ points. Although the data shown in Tables 1 and 2 are similar in most respects, results differ with respect to the overall magnitude of the Flynn effect (14.0 vs. 7.5 IQ points for WAISR and WAISIII, respectively). Part of this difference would be expected on the basis of Flynns (2009) determination that the WAISIII yields a Full Scale IQ that is at least 1.65 points higher than the correct value. The anomaly in the WAISIII norms has been difficult to quantify precisely, but Flynn regards the 1.65-point correction as a conservative adjustment. When we adjusted the WAISIII subtest means by 1.65 IQ points (or 0.33 scaled-score unit), the estimated magnitude of the overall Flynn effect for the WAISIII becomes 12.3 IQ points over a 50-year interval, which is much closer to our estimate for the WAISR. In addition, differences between the validity samples for the WAISR and WAISIII may have contributed systematic error to the estimation of the magnitude Flynn effect for each subtest. As pointed out by Zhu and Tulsky (1999), these validity samples, from which our Flynn effect estimates are derived, should not be confused with the larger and more carefully selected standardization samples. Nevertheless, the TAE means obtained for verbal and performance WAISIII subtests are similar to corresponding WAISR means, and the overall means are identical.

868

DICKINSON AND HISCOCK

EXPERIMENT 3 Method Our final calculations combine data from the WAISR and WAISIII to generate an optimized model of the respective contributions of age group and the Flynn effect on each of the 11 Wechsler subtests. Unweighted means representing change over 50 years were calculated and expressed as scaled-score units. FED calculations were made with and without adjustment for the previously described 1.65-IQ-point (0.33 scaled-score unit) anomaly in the WAISIII norms. Results and conclusions

Downloaded by [Universitara M Emineescu Iasi] at 05:17 17 November 2011

The results are shown in Figure 2. The overall 50-year AGD is 2.6 scaled-score units (13.0 IQ points), and the comparable FED, after adjustment for the putative anomaly in WAISIII norms, is also 2.6 scaled-score units (13.0 IQ points). Taken at face value, the calculations indicate that the Flynn effect accounts for 100% of the difference between 20- and 70-year-olds. The mean TAE is zero.

Results for verbal and performance subtests are dissimilar. Data for the verbal subtests, depicted in the upper panel of Figure 2, indicate that the magnitude of the AGD is 0.8 scaled-score unit (4.0 IQ points), and the magnitude of the FED is 2.4 scaled-score units (12.0 IQ points). By subtraction, the verbal TAE is +1.6 scaled-score units, or +8.0 IQ points. After adjustment for the Flynn effect, scaled scores for 70year-olds on all verbal subtests except Digit Span are higher than the average for 20-year-olds. For performance subtests, the magnitude of the AGD is 4.9 scaledscore units (24.5 IQ points), and that of the FED is 3.0 scaled-score units (15.0 IQ points.) The resultant TAE is 1.9 scaled-score units, or 9.5 IQ points. Scores on performance subtests remain substantially below the average for 20-year-olds after adjustment for the Flynn effect. As shown in Figure 2, the results of the Flynn effect are attenuated somewhat if the 1.65-IQ-point adjustment of WAISIII scores is not made. The overall 50year FED is reduced to 2.2 scaled-score units (11.0 IQ points), and the mean TAE becomes 0.4 scaledscore units (2.0 IQ points). The Flynn effect in this instance accounts for 85% of the difference between 20- and 70-year-olds.

Verbal Subtests
14 12 Mean Scaled Score 10 70-year-olds 8 6 4 2 0 Inform Dig Span Vocab Arith Compr Simil 70-year-olds (FE adjustment) 70-year-olds (FE and WAIS-III adjustment) 20-year-olds

Performance Subtests
12 20-year-olds 10 Mean Scaled Score 8 6 4 2 0 Pic Comp Pic Arr Block Dgn Obj Assy Dig Sym
Figure 2. Mean scaled scores for 20-year-olds and for 70-year-olds with and without adjustments for the Flynn effect and the anomaly in Wechsler Adult Intelligence ScaleThird Edition (WAISIII) norms identified by Flynn (2009). FE = Flynn effect.

70-year-olds

70-year-olds (FE adjustment) 70-year-olds (FE and WAIS-III adjustment)

AGE-RELATED IQ DECLINE

869

Downloaded by [Universitara M Emineescu Iasi] at 05:17 17 November 2011

Irrespective of whether Flynns suggested correction is applied to the WAISIII norms, the estimated overall magnitude of the Flynn effect falls below the 3 points per decade figure that has been reported by Flynn (1984, 1987, 1999, 2006). Our estimate of 2.4 IQ points per decade for verbal subtests (with WAISIII adjustment) approximates the value of 2 points cited by Flynn but the corresponding estimate of 3.0 points per decade for performance tests falls below Flynns figure of 4 points. The apparent underestimation of the Flynn effect on performance scores is observed in both the change between WAIS and WAISR (Table 1) and the change between WAISR and WAISIII (Table 2). The discrepancy presumably can be attributed to our employment of a computational procedure that differs from the standard procedure for determining IQ. In any event, the reduced magnitude of the Flynn effect in our calculations implies that we may have underestimated the actual impact of the Flynn effect on age differences.

GENERAL DISCUSSION Older people, in principle, are doubly disadvantaged when faced with many of the tasks that constitute an IQ test. First, as a consequence of the aging process, certain intrinsic abilities such as working memory and speed of cognitive processing may have diminished (e.g., Craik & Salthouse, 2007). Until recently, explanations for lower performance in older adults have focused on those kinds of intrinsic factors. However, it is now recognized that, irrespective of any cognitive deterioration during their lives, older individuals are disadvantaged in addition by their birth datethat is, by the Flynn effect. Population averages have risen with the addition of each new cohort, and test norms have risen accordingly. Even when they were 20 years old, todays 70-year-olds would have scored below the level of todays 20-year-olds. Given that both intrinsic aging and birth date contribute conjointly to the lower performance of older people on Wechsler IQ subtests, it follows that adjusting for either of those factors will reduce differences between older and younger groups. Our findings confirm that adjusting for the Flynn effect reduces the performance differential between 20- and 70-year-olds. The more notable finding is the magnitude of the required adjustment. Our best estimate, based on pooled data from the WAISR and WAISIII, is that at least 85% of the overall difference between the age groups is attributable to the Flynn effect. If one uses Flynns (2009) suggested correction to WAISIII norms, which he regards as conservative, the estimate increases to 100%. Our estimates are even higher than Hiscocks (2007) calculation that the Flynn effect accounts for 71% of the age-related decline in raw scores on Ravens Matrices test. Moreover, these high estimates are obtained even though our calculations for the Wechsler IQ tests underestimate the overall strength of the Flynn effect as measured by other means (Flynn, 2009). The data indicate an enormous contribution of the Flynn effect to age differences in

performance on IQ tests, and this will require modifications in the way in which age-related performance declines are interpreted. Two implications are especially salient. The first concerns verbal tests. Although the Flynn effect depresses both verbal and performance scores for older groups in the cross-sectional norms, the verbal scores (other than Digit Span) do not decline as much as they would be expected to decline on the basis of the Flynn effect alone. Accordingly, the skills measured by these five verbal subtests must actually improve between the ages of 20 and 70 years. Longitudinal studies, most of which have used cognitive measures other than IQ tests, support the conclusion that verbal skills reach their peak between the ages of 50 and 70 years, depending on the specific skill, and remain relatively intact beyond the age of 70 (Ghisletta & Lindenberger, 2004; Park et al., 2002; Schaie, 1994). Thus, with respect to verbal abilities, adjusting cross-sectional IQ norms for the Flynn effect brings those norms into agreement with longitudinal studies of cognitive aging. The second implication is that the Flynn effect must be considered before interpreting an abnormal degree of intellectual decline as accelerated aging (Bartus, 1980). As much as two thirds of the normative decline in raw scores on the five performance subtests plus Digit Span can be attributed to a cohort effect and not to a decrease in ability over time. It may seem paradoxical, but those six subtests are much less sensitive to the effects of aging than the cross-sectional norms would indicate. Thus, even moderate deterioration of performance on one of those subtests, relative to the patients age group, is numerically equivalent to a very large number of years of normal aging. It might be advisable to regard such deterioration as a frank indication of abnormality rather than to invoke the concept of accelerated aging. The present study suggests a new methodology for investigating age-related change in various cognitive abilities. Cohort effects notwithstanding, the crosssectional design may be the only method available when circumstances do not allow a longitudinal study to be done. The procedure that we describe here, which adjusts cross-sectional age differences for the Flynn effect, constitutes a means of increasing the validity of cross-sectional comparisons. The same rationale and methodology can be applied to age-related cognitive changes other than IQ changes, provided that it is possible to calculate the degree to which scores in the general population have changed across cohorts and provided that the rate of change is relatively constant over a period of many years. Even if the magnitude of the Flynn effect cannot be specified with a high degree of accuracy, adjustments based on reasonable estimates will yield more valid results than no adjustments at all (Flynn, 2009). A final note concerns the Wechsler Adult Intelligence ScaleFourth Edition (WAISIV; Wechsler, 2008), which was standardized in 2006, or about 11 years after standardization of the WAISIII. Comparison of the respective norms indicates that the Flynn effect as reflected in Full Scale IQ has continued unabated during that 11-year interval (Flynn, 2009). However, various

870

DICKINSON AND HISCOCK Flynn, J. R. (1999). Searching for justice: The discovery of IQ gains over time. American Psychologist, 54, 520. Flynn, J. R. (2006). Tethering the elephant: Capital cases, IQ, and the Flynn effect. Psychology, Public Policy, and Law, 12, 170189. Flynn, J. R. (2007). What is intelligence? Beyond the Flynn effect. Cambridge, UK: Cambridge University Press. Flynn, J. R. (2009). The WAIS-III and WAIS-IV: Daubert motions favor the certainly false over the approximately true. Applied Neuropsychology, 16, 98104. Ghisletta, P., & Lindenberger, U. (2004). Static and dynamic longitudinal structural analyses of cognitive changes in old age. Gerontology, 50, 1216. Hiscock, M. (2007). The Flynn effect and its relevance to neuropsychology. Journal of Clinical and Experimental Neuropsychology, 29, 514529. Neisser, U. ( Ed.). (1998). The rising curve: Long-term gains in IQ and related measures. Washington, DC: American Psychological Association. Park, D. C., Lautenschlager, G., Hedden, T., Davidson, N. S., Smith, A. D., & Smith, P. K. (2002). Models of visuospatial and verbal memory across the adult life span. Psychology & Aging, 17, 299320. Schaie, K. W. (1986). Beyond calendar definitions of age, time and cohort: The general developmental model revisited. Developmental Review, 6, 252277. Schaie, K. W. (1994). The course of adult intellectual development. American Psychologist, 49, 304313. Tulsky, D., Zhu, J., & Ledbetter, M. ( Eds.). (1997). WAIS-III WMS-III technical manual (3rd ed.). San Antonio, TX: Harcourt Brace. Wechsler, D. (1955). The Wechsler Adult Intelligence Scale. New York: The Psychological Corporation. Wechsler, D. (1981). The Wechsler Adult Intelligence Scale Revised. New York: The Psychological Corporation. Wechsler, D. (1997). The Wechsler Adult Intelligence ScaleThird Edition. San Antonio, TX: The Psychological Corporation. Wechsler, D. (2008). The Wechsler Adult Intelligence ScaleFourth Edition. San Antonio, TX: The Psychological Corporation. Zhu, J., & Tulsky, D. S. (1999). Can IQ gain be accurately quantified by a simple difference formula? Perceptual and Motor Skills, 88, 12551260.

changes in the content, administration procedures, and scoring rules of the WAISIV preclude extending our analyses of the original 11 WAIS subtests to the WAISIV. If most of the characteristics of the WAISIV should be preserved in a subsequent revision, then it will be feasible to apply our method to normative data from the WAISIV and its successor. Such an analysis would be completely independent of the present analyses and thus would be capable of providing convincing confirmation (or refutation) of the present conclusions.
Original manuscript received 17 August 2009 Revised manuscript accepted 4 January 2010 First published online 26 March 2010

Downloaded by [Universitara M Emineescu Iasi] at 05:17 17 November 2011

REFERENCES
Bartus, R. T. (1980). Cholinergic drug effects on memory and cognition in animals. In L. W. Poon ( Ed.), Aging in the 1980s ( pp. 163180). Washington, DC: American Psychological Association. Craik, F. I. M., & Salthouse, T. A. ( Eds.). (2007). The handbook of aging and cognition (3rd ed.). Hove, UK: Psychology Press. Dickens, W. T., & Flynn, J. R. (2001). Heritability estimates versus large environmental effects: The IQ paradox resolved. Psychological Review, 108, 346369. Flynn, J. R. (1984). The mean IQ of Americans: Massive gains 1932 to 1978. Psychological Bulletin, 95, 2951. Flynn, J. R. (1987). Massive IQ gains in 14 nations: What IQ tests really measure. Psychological Bulletin, 101, 171191. Flynn, J. R. (1998a). IQ gains over time: Toward finding the causes. In U. Neisser ( Ed.), The rising curve: Long-term gains in IQ and related measures ( pp. 2566). Washington, DC: American Psychological Association. Flynn, J. R. (1998b). WAIS-III and WISC-III IQ gains in the United States from 1972 to 1995: How to compensate for obsolete norms. Perceptual and Motor Skills, 86, 12311239.