Tripp A, Tollefson N
J Nurs Educ. 1985 Mar;24(3):92-8. doi: 10.3928/0148-4834-19850301-04.
Differences in item difficulty between conventional and complex multiple-choice (M-C) items measuring objectives written at the six levels of the cognitive domain and three content areas within a basic health professions subject matter area were investigated. Conventional and complex M-C items were also compared in terms of item discrimination. Five hundred fifteen senior baccalaureate nursing students randomly responded to one of two packages of 18 one-item pharmacology achievement tests. Items in the two packages were matched in all respects except for the type of option. A 2 X 3 X 6 repeated measures ANCOVA with method of pharmacology as the covariate showed a statistically significant (p less than .01) interaction of option X content X cognitive level. Scheffé tests revealed that complex items were more difficult than conventional option items at the knowledge level only across all content areas investigated. No statistically significant differences were found between the two option types discriminating ability. Results of this research question the validity of the assumption that complex items are more difficult and discriminating than conventional M-C items.
