The use of open-end data as an aid in writing multiple-choice distracters: An evaluation with arithmetic reasoning and computation items.

Data gathered during the development of the Navy Arithmetic Test were analyzed to evaluate the utility of using distracters derived by administering the items in open-ended (OE) form. For both Computation items and Reasoning items moderately high correlations (about .50-.60) were found between frequency with which responses were written in by Ss in OE format and frequency with which the same responses were chosen in multiple-choice format. That OE-derived responses tend to retain their relative popularity in multiple-choice format appears to provide some support for the use of the relatively expensive OE technique in arithmetic test construction. Caution is urged in applying these findings to other test types. From Psyc Abstracts 36:05:5KJ31R. (PsycINFO Database Record (c) 2010 APA, all rights reserved)