NotesFAQContact Us
Search Tips
ERIC Number: ED173354
Record Type: Non-Journal
Publication Date: 1978
Pages: 11
Abstractor: N/A
Reference Count: N/A
Q. How Many Options Should a Multiple-Choice Question Have? (a) 2. (b) 3. (c) 4. At-a-glance Research Report.
Catts, Ralph
The reliability of multiple choice tests--containing different numbers of response options--was investigated for 260 students enrolled in technical college economics courses. Four test forms, constructed from previously used four-option items, were administered, consisting of (1) 60 two-option items--two distractors randomly discarded; (2) 40 three-option items--one distractor randomly discarded; (3) 40 three-option items--the least appropriate distractor discarded; and (4) 30 of the original four-option items. Data from students who didn't complete the test or who omitted many items were disregarded. Results indicated that three-option items were as reliable as four-option items (reliabilities were .6 to .7 and .7, respectively); that higher scores were obtained on tests with fewer options per item; and that tests consisting of four-option item required more time for both test construction and administration. A comparison of tests (2) and (3), both containing three-option items, indicated that test (2) was more reliable, less difficult, and took longer to complete. (GDC)
Publication Type: Reports - Research
Education Level: N/A
Audience: N/A
Language: English
Sponsor: N/A
Authoring Institution: New South Wales Dept. of Education, Sydney (Australia).
Identifiers - Location: Australia