ERIC Number: ED255580
Record Type: Non-Journal
Publication Date: 1985-Apr
A Quantitative Review of Research on Multiple-Choice Item Writing.
Haladyna, Thomas M.; Downing, Steven M.
In this paper 45 item-writing rules for multiple-choice tests presented in textbooks on educational measurement in a previous study are identified. The current study presents a quantitative review of the literature with respect to the empirical and theoretical evaluation of these principles of item-writing. Fifty-six studies that addressed at least one of the 45 item-writing rules were identified. Twenty-one of the rules have been studied empirically; 24 item-writing rules have no empirical basis. The optimal number of options was the most frequently studied rule, with 18 studies cited. The major generalization from these studies is that three options maximize test reliability and efficiency. Type-k items were evaluated in eight studies. Results suggest that compared to single-answer multiple-choice items, type-k items are more difficult, provide clues to some examinees, and decrease test reliability and efficiency. Eight other studies suggest some empirical basis for keeping the length of the keyed option about the same as other options. This review suggests that the majority of the common principles of multiple-choice item writing are not empirically based. (Author/DWH)
Publication Type: Speeches/Meeting Papers; Information Analyses; Reports - Research
Education Level: N/A
Sponsor: American Coll. Testing Program, Iowa City, IA. Research and Development Div.
Authoring Institution: N/A