Nonfunctioning Options: A Closer Look

Abstract

In a previous study, Cizek and O'Day reported the effects of eliminating nonfunctioning options from five-option, multiple-choice test items. That study, however, only investigated the effect of option removal on item performance when items were studied as a set and not the effects of option removal for individual items. The authors also did not examine potential changes in dimensionality resulting from altering the items. The present study provides a closer look at the effect of removing nonfunctioning options by examining changes in difficulty, discrimination, and dimensionality. Results of these analyses supported the conclusions of the earlier research and provide additional support for benefits that may accrue from eliminating nonfunctioning options: enhanced score reliability, reduced testing time, potential for broader domain sampling, and consequent increased validity of score interpretation.