A Comparison of Linear, Fixed Common Item, and Concurrent Parameter Estimation Equating Procedures in Capturing Academic Growth

A natural method to assess academic growth, and in turn facilitate inferences regarding the effectiveness of educational policy, administration, and instruction, is to compare the performance of students on standardized tests over time. A key methodological step in this process is equating test scores from separate administrations so that the scores are comparable. Three common item response theory equating approaches are investigated: linear transformation of separate calibrations, fixed common item parameter calibration, and concurrent calibration. Using data from a high-stakes statewide testing program, differences in mean growth were found between methods depending upon both the type of ability estimate used and the type of equating procedure implemented. Furthermore, depending upon the choice of equating method, large differences in classification may occur.