A Longitudinal, Criterion-Related Validity Study of the Cognitive Levels Test for a Group of Highly Gifted Students

Abstract

Using a postdictive design, the criteron-related validity of the Cognitive Levels Test(CLT) was evaluated for a group of 48 highly gifted 10-year-old students. At the time of original testing on the Stanford-Binet Intelligence Scale (SB), the children averaged 3 years, 8 months of age. After a 6-year, 4-month interval, the median correlation between the SB and CLT standard scores was .46. When corrected for curtailment, the median correlation was. 71. Given the 6.5-year test-retest reliability of the SB for a sample of normal children (r = .66), the correlation between the CLT and SB for a restricted sample of gifted children was taken as strongly supportive of the validity of the CLT as a measure of cognitive ability. After a 14-month interval, the median correlation between the standard scores of the CLT and California Achievement Test Total was .50. Following correction for curtailment, the median correlation was .64. Therefore, the estimation of academic achievement by the CLT was also considered as supportive of the CLT, particularly in light of the restricted range of ability of the present sample.