Generalizability of Gains on State Exams
PIs: Paul Peterson, Harvard University; William Howell, University of Chicago

Summary:
What is the relationship of accountability to school and student performance?This project is analyzing state exams in terms of performance levels and annual gains.  Specifically, it is looking at the individual, sub-group, and across-time correlations between performance on the criterion-referenced and norm-referenced tests. The Florida Department of Education provided data for this research, which includes test scores and demographic information on each public school student in the state for the school years 1998-99 through 2002-03.

Findings:
Earlier work examining the Florida A+ Accountability Plan indicates that its use of “A” through “F” distinctions creates a more accurate measuring stick for school progress than the federal NCLB’s use of the “adequate yearly progress” designation, which distinguishes between schools on that basis alone. Researchers on this project hold that the NCLB method of classifying schools is both opaque and legalistic, obscuring as much as it reveals. And, unlike the Florida A+ Plan, it fails to directly measure student growth and therefore cannot discriminate clearly between higher quality and lower quality schools. Further research on state exams is ongoing.


Related Publications