Impact of seasonality in data and calibration timing on Rasch parameters

• In Event: IRT calibrations

Abstract

When seasonality exists in the data, and the data collected during field testing period contains student responses from only a particular time window, item estimates may be biased. This research investigates seasonality effects and their impacts on item parameters and scaled scores using variable-length CAT data.

Authors

• Athul Sudheesh

• Sonya Powers, Edmentum, Inc.

• Yi He, Edmentum