Paper Summary
Share...

Direct link:

Prediction of Student Proficiency on High-Stakes Assessment From Results of Benchmark Assessments

Tue, April 17, 2:15 to 3:45pm, Vancouver Convention Centre, Floor: Second Level, East Room 12

Abstract

The present study examined the predictive accuracy of several methods to predict student proficiency on the state high stakes assessment (PSSA) from the result of benchmark assessments administered in grades 3 through 5 in Pittsburgh Public Schools. The outcomes showed that Logistic Regression analysis provided the best accuracy in predicting percentage of proficient students on the PSSA in schools. Linear regression analysis has been found to be inaccurate in most cases to find optimal cut scores to predict student proficiency on the PSSA. Suggestions are made to use a simple version of Logistic Regression analysis to estimate student proficiency probabilities on the PSSA for new editions of benchmark assessments.

Authors