Abstract
The purpose of this study was to investigate the effects of multiple-choice item stem completeness and orientation on item difficulty and discrimination. Six hypotheses were tested using archival examinee data from the reading comprehension sections of 30 employment selection exams. Three multivariate analyses revealed that item difficulty and discrimination were not significantly affected (p > .05) by stem completeness or orientation. The Flesch-Kincaid Grade Level of the reading passage, passage length, stem length, and key length were used as covariates in the first two analyses, while no covariates were used in a third exploratory analysis. Limitations and theoretical implications of the findings were also discussed.