Return to search

An investigation of test bias of a kindergarten screening battery in predicting achievement and educational placement for American Indians and Caucasians

The purpose of this study was to examine whether the relationship between a kindergarten screening battery (KSB) and two criterion variables (academic achievement and educational placement) varied as a function of race. The subjects were 702 elementary school students, approximately 176 per grade, in kindergarten, first, third, and fourth grade. Approximately 45 students per grade were American Indian, with the remainder Caucasian.Multiple regression analysis was used to examine the relationship between the KSB, race, and its interaction with each criterion variable. The predictor variables were the Peabody Picture Vocabulary Test-Revised (PPVT-R), the Kindergarten Language Screening Test (KLST), the Developmental Visual-Motor Integration Test (VMI), the Draw-A-Child (DAC), race, and the four race x KSB test vectors. Criterion variables were the Stanford Achievement Test (SAT) total battery, and educational placement (special education vs. regular education status).The KSB predictor variables were entered into the multiple regression equations as a block. Race was then entered as a coded group vector, followed by the four race x KSB test interaction vectors. The increment in R2 due uniquely to the interaction was nonsignificant (slopes were homogeneous) in all analyses. Both the effects of race and the KSB were significant for all grades with achievement as the criterion (p < .01).Bias in predicting educational placement was tested using all subjects combined. Both the interaction and race effects were nonsignificant. The main effect of the KSB was significant (p < .01). An improvement over chance analysis showed that the KSB increased predictive accuracy of the at-risk determination over chance.The weighted composites which best predicted achievement consisted primarily of the language instrument, the PPVT-R, and race. The two psychomotor tests (the DAC and VMI) contributed useful information at the two earlier grades. The composite which best predicted educational placement was made up primarily of the expressive language measure, the KLST.Results of the study supported continued use of the KSB. However, race should be used as a variable when predicting achievement from the KSB, so as not to overpredict American Indian achievement or underpredict Caucasian achievement. Results also indicated the potential utility of early language enrichment for improving achievement of high risk children. / Department of Educational Psychology

Identiferoai:union.ndltd.org:BSU/oai:cardinalscholar.bsu.edu:handle/181199
Date January 1988
CreatorsStone, Brian J.
ContributorsGridley, Betty E.
Source SetsBall State University
Detected LanguageEnglish
Formatxi, 137 leaves ; 28 cm.
SourceVirtual Press
Coveragen-us---

Page generated in 0.0021 seconds