Department of Psychology
In this study l/the researcher sought to examine the correlational and classificational agreement (sensitivity and specificity) between two very popular screening instruments which have undergone recent revisions-The Developmental Indicators for the Assessment of Learning-Revised (DIAL-R) and the Denver Developmental Screening Test-ll (Denver-ll) and one more recently published new screening instrument on the market--The Early Screening Profiles (ESP). The sample for this study consisted of 60 preschool children attending two federally and state funded preschool programs in Western Kentucky. The children were of the correct age to enter kindergarten in the fall of 1993. Results of this study revealed that the three instruments themselves showed an inability to classify children similarly. Using the DIAL-R as a criterion, the ESP and Denver-ll showed high specificity (88.9%-100%) but low sensitivity (25%-30%). While the correlation between the DIAL-R and ESP scores was moderately high (.72), its classification agreement was discrepant. One further interesting finding was that the group mean scores for children were significantly higher on the DIAL-R than on the ESP (mean difference=9.74 points). As the true job of any screening instrument is to accurately identify children who fall into high risk categories, these instruments appear to be "missing" many children. Analyzing all three instruments' classification agreements together resulted in discrepant identification of children.
Education | Psychology
Batchelor, Shannon, "The Concurrent Validity of Three Preschool Screening Instruments" (1994). Masters Theses & Specialist Projects. Paper 928.