ERIC Number: ED064388
Record Type: RIE
Publication Date: 1972-Apr
Reference Count: 0
Comparative Reliabilities and Validities of True-False and Multiple Choice Tests.
Frisbie, David A.; Ebel, Robert L.
A study designed to compare the reliabilities of multiple choice and true-false tests that were constructed to measure the same objectives was conducted. The impetus for this study came from the research reported by Ebel (1971) on the same topic. Subjects were selected from six public high schools. Three phases of testing were required for instrument development and data gathering. Phase I involved collecting item analysis data for one item conversion method and Phase II was used to try out the true-false items. The final phase of testing included 1018 students responding to eight final test forms. The social studies and natural science multiple choice items employed in this study appeared in a widely used battery of achievement tests. The original 70-item multiple choice tests SM (social studies) and NM (natural science) were each administered to a minimum of 100 subjects. The four true-false test forms were each administered to a minimum of 50 subjects in Phase II. The eight final test forms varied according to subject matter, item conversion method, and item form order. The results of this study support the notion that students respond to more true-false than multiple choice items in a given period of time. However, the data indicate that the multiple choice tests were more reliable though they tended to measure the same thing that the true-false tests measured. (CK)
Publication Type: N/A
Education Level: N/A
Authoring Institution: N/A
Note: Paper presented at the annual meeting of AERA (56th, Chicago, Ill., April 3-7, 1972)