A longitudinal predictive validity investigation of the SB:FE and K-ABC with at-risk children

The predictive validity of the Stanford-Binet Intelligence Scale: Fourth Edition (SB:FE) and Kaufman Assessment Battery for Children (K-ABC) was investigated by testing a sample of African-American and White children from impoverished families over an 8-year period. The cognitive tests were administered at ages 4, 6, and 9, and the Metropolitan Achievement Test was administered each year from kindergarten through sixth grade. Predictive validity coefficients for both tests were moderately strong and quite consistent over time. The correlations between the cognitive and achievement measures were adequate for the youngest group of children and improved when the predictors were administered at later ages. The K-ABC Achievement scale was found to he the best predictor of academic achievement, particularly for the school-aged children. Each test predicted academic performance of both African-American and White children equally well.