Research report

A comparison of item parameter estimates in Pychometrik and the existing item calibration tool


By: Wei He, Patrick Meyer, Russ Grandgeorge, Emily Bo


The design features of the existing MAP Growth item calibration system have become too restrictive to support the expanding business and product development needs of NWEA. As such, Pychometrik, a new item calibration tool built with Python, was recently developed to allow psychometricians autonomy in the item calibration process for all NWEA shelf products. To evaluate the performance of Pychometrik against the existing tool, two studies were conducted: (1) an item parameter estimate comparability study using both real data and simulations and (2) an item parameter recovery simulation study to evaluate the parameter recovery accuracy of the new tool. Results from both studies support the use of Pychometrik for future item calibrations.

See More
View Asset

Related Topics