Measurement error is a common theme in classical measurement models used in testing and assessment. In classical measurement models, the definition of measurement error and the subsequent reliability coefficients differ on the basis of the test administration design. Internal consistency reliability specifies error due primarily to poor item…

Outlines a method to write multiple (four) versions of each examination. Reviews some advantages and disadvantages of multiple-choice and open-ended problems. Considers the writing and grading techniques used. Reports that the examination met the consistency requirement and received a good evaluation from students. (YP)