Multiple Choice and True-False

Publisher Website
Google Scholar
Abstract
Methodological problems with previous comparison of multiple-choice (MC) and true-false (TF) tests include differential item improvement, nonequivalence of stimuli, and restricted range for the TF form. This paper compares reliabilities and concurrent validities of tests composed of unimproved, teacher-made items. Form equivalence was assured by using every MC option as a TF item; range was considered comparatively unrestricted since the TF test took longer to complete than the MC test. No significant differences were found in either reliability or validity even when the MC reliability was adjusted to equate testing time. These results provide support for the use of TF tests as alternatives to MC tests in objective testing and are discussed in relation to previous findings.