| Summary: | The early identification of students at-risk for future learning problems
typically forms the basis for the Implementation of early intervention programs
designed to prevent, diminish, and/or correct learning difficulties. Kindergarten
screening results influence the allocation of special services and are linked with the
expenditure of monetary and personnel resources. The extent to which screening
contributes to accurate and useful educational decision-making requires evaluation.
The purpose of the present study was to investigate the validity and utility
of four kindergarten screening measures and their composite screening
classification as predictors of third grade achievement. History of school-based
intervention and retention status were considered to be additional indices of school
performance and their relationship with kindergarten screening results was also
investigated. The screening measures included the Draw-A-Person, the
Kindergarten Language Screening Test, the Mann-Suiter Visual Motor Screen, and
the Deverell Test of Letters and Numbers. The achievement measure employed
was the Canadian Tests of Basic Skills.
Validity data indicating the degree of accuracy of screening classification
decisions (risk/no-risk) was possible through the utilization of decision matrix
analysis. Interpretations in this study included percentage calculations of the
problem base rate, referral rate, and overall hit rate. Vertical evaluation presents
prediction accuracy in relation to criterion (actual) performance versus horizontal
evaluation which is calculated in relation to screening (predicted) performance.
Prediction-performance matrices presented in this study represent data
available for one age coliort of 684 subjects enrolled since kindergarten in one
school district located near Vancouver, British Columbia. Seven achieved samples
were generated, the number of subjects ranging from 576-663.
The results of this study demonstrate that screening referral rates were less
than their respective problem base rates, indicating general under-referral of at-risk
students. For all analyses, vertical evaluation more appropriately demonstrated
greater under-referral rates than did horizontal evaluation. Vertical evaluation also
contributed to greater accuracy of interpretation than did horizontal evaluation
which proved to be misleading. Specificity rates (vertically calculated true
negatives) were much larger than sensitivity rates (vertically calculated true
positives), indicating far greater accuracy for the identification of non-risk than
at-risk students. Overall hit rates were high but misleading as the proportions of
correctly identified at-risk and non-risk students were not indicated. === Education, Faculty of === Educational and Counselling Psychology, and Special Education (ECPS), Department of === Graduate
|
|---|
