Description of variability of subtest scatter across cognitive bands on the Test of Cognitive Skills (TCS(RTM))

Date of Completion

January 2004


Education, Educational Psychology|Psychology, Psychometrics|Psychology, Cognitive|Education, Curriculum and Instruction




Several issues surround the use of IQ and subtest scores. Two controversies are that identical IQs estimate the same amount of g and represent a homogeneous group. Another is that, although the full-scale score shows predictive validity for achievement, the rigor of the subtest scores are weak and should not be used to describe strengths and weaknesses. Aspects of these issues will be addressed in this study. ^ The purpose of this exploratory study was to describe the nature of variability within individuals across cognitive bands using the Test of Cognitive Skills (TCS®/2). The scores of two cohorts, a fall sample of 1,226 students and a spring sample of 707, were divided into cognitive bands of five points, from 58 to 141, except for the lowest and highest bands. The discrepancies scores of all pair-wise comparisons (Sequence, Analogies, Memory, Verbal Reasoning, Non-Verbal factor, high/low scores) were calculated, and their relationships to the total score examined. The results of the high-low scores were compared to those found with the WISC-R, WISC-III, WAIS, WPPSI, and WJ-R. ^ The results indicated that subtest scatter was prevalent across all cognitive bands, much larger than what the Standard Error of Measure could explain. By using the averages of the cognitive bands, the shared variance became 42% to 84%. Significant differences were found between several of the abilities and factors. The slope of Memory was different from the other cognitive abilities of Sequences, Analogies, and Verbal Reasoning. Two learning profiles were identified. Verbal Reasoning was the high score for a majority of gifted, whereas Memory was the high score for those at the lower end of the continuum. A transition occurred from one end of the continuum to the other, and a developmental change was also evident. ^ In comparing the TCS®/2 with the WISC-R, WISC-III, WPPSI, WAIS, and WJ-R , no significant differences were found. The TCS ®/2 showed the same characteristics for high/low scores as the individual intelligence tests. Thus, a group intelligence test can describe significant differences between cognitive bands, and the results provided evidence that identical IQ scores may not represent homogeneous groups. ^