The Brigance K&1 Screen: Factor Composition with a Head Start Sample

Abstract

This study examined the factorial structure of the Brigance K&1 screen with a sample of 383 Head Start children from a Midwestern community. Maximum likelihood estimation via LISREL 8 was used to conduct confirmatory factor analyses using the covariance matrix of age-based linear standard scores on each of the 12 Brigance sub-tests. The factors reported in the Brigance technical report, which serves as the model on which Brigance domain scores are obtained, did not fit the independent set of data. Rather, a competing three-factor model, based on 7 of the 12 Brigance subtests, measuring Rote Knowledge (Factor 1), Visual-Spatial Skills (Factor 2), and Comprehension (Factor 3), showed excellent fit to the data. Concurrent validity evidence indicated low to moderate associations when each factor was correlated with the cognitive processing subtests of the K-ABC. The predictive validity of the factors was investigated using teacher ratings, performance on standardized achievement tests, and special education placement. Predictive validity coefficients ranged from low to moderate. The results bring into question the use of this test with the Head Start population.