The 'Flynn Effect': Does it apply to academic achievement?

Abstract

Recent years have witnessed extensive interest in Flynn's research findings, which have been subsumed under the rubric (FE). This research, conducted in 14 countries, has revealed a virtually invariable tendency for IQ scores to progressively increase. Therefore when an aptitude test was renormed, an examinee had to secure a higher raw score to keep pace with the earlier standard score. However, Flynn has questioned how aptitude can rise if people do not learn easier and if they fail to experience higher levels of occupational success. The present inquiry employed several versions of the Wide Range Achievement Test (WRAT) to assess whether the FE prevails when an achievement test is renormed. Results on the Arithmetic subtest affirmed the FE, whereas results on the Reading and Spelling subtests did not. Replication studies are suggested, to appraise the implications of Gc and Gf, as well as maturation, on longitudinal aptitude and achievement trends. Key Words: IQ, The Flynn effect, Academic Achievment Directionality Americans place great importance on education, and presently confront a puzzling inconsistency: how it is possible for longitudinal studies to report that U.S. students perform lower on academic achievement tests (Williams & Ceci, 1997), but obtain progressively higher IQ scores (Begley, 1996; Flynn, 1996; Loehlin, 1997; Hall, 1998)? Long-term and upward trajectories of aptitude profiles have been documented by James R. Flynn, a political philosopher at the University of Otago, New Zealand. Flynn examined and assessed longitudinal aptitude test trends by decade in 14 nations (Flynn, 1984, 1987, 1996). Through a series of investigations, Flynn observed that every time an aptitude test was renormed, an examinee had to secure a higher raw score in order to keep pace with the standard score he/she had previously obtained (Kamphaus, 1993, p. 120): this progressive rise in IQ over decades has been termed the Flynn Effect (FE). Should results of FE investigations prove accurate, they raise potentially significant social and political implications. Some studies indicate that IQ score increments are largely concentrated in the lower social economic groups (Jensen, 1998, pp 319-320). This reduced the standard deviation and may in effect mask a decline in scores of the middle and upper classes. While widely cited, aptitude data supporting the FE are mysteriously erratic; the per-decade increment in standard scores ranged from less than two standard points in Great Britain to more than 12 standard score points in East Germany. Such an erratic range calls into question the validity of FE investigations. While accepting FE as a real phenomonon, Jensen 1998 p.319) notes that test profiles vary by test and time period, and that the variables across studies are utterly confounded. In the 14 nations wherein data were gathered, IQ gains extended from 5 to 25 points in one generation (Flynn, 1987). A primary problem rests with the samplig issues. A large proportion of FE studies employ tests which are normed based on representative population at one point, and renormed many years later on a different but presumably similar population. For these studies, sample equivalences cannot be assumed (Jensen, 1998, p. 318). Still another proble with the FE involves practical significance. As Jensen notes, one correct item on the Ravens Matrices alone can increase a subject's score by two to three points. Significantly,ca large proportion of FE studiee employed the Ravens Matrices. Assuming that the FE is an authentic and empirically validated phenomenon, implications are unclear. Flynn has acknowledged difficulty in reconciling progressively higher IQs and the absence of a corresponding increase in the general population of what is considered practical intelligence. Thus he notes that more people do not find school easy, and occupational success has not been more apparent during the decades in which IQ increases have been documented (Flynn, 1987, p. …