Author replies to article TM 500 559. (MB)

The procedure of estimating true scores by means of a transformation of the obtained score based on the reliability coefficient is compared with the use of the obtained score without transformation. Using the mean squared error as a criterion, the transformed score is a better estimate for most examinees but poorer for those whose true scores lie…

When test-takers are offered a choice of essay questions, some questions may be harder than others. If the test includes a common portion taken by all test-takers, an adjustment to the scores is possible. Previously proposed adjustment procedures disregard the test-makers' efforts to create questions of equal difficulty; these procedures tend to…

A distinction is made between reliability of measurement and reliability of classification; the "criterion-referenced reliability coefficient" describes the former. Application of this coefficient to the probability distribution of possible scores for a single student yields a meaningful way to describe the reliability of a single score. (Author)

This article is a reply to a previous paper (see TM 500 488) interpreting Livingston's original article (see TM 500 487). (CK)

A reliability coefficient for criterion-referenced tests is developed from the assumptions of classical test theory. The coefficient is based on deviations of scores from the criterion score, rather than from the mean. (Author/CK)

Practical methods of setting standards for the Basic Skills Assessment tests in secondary level reading, writer's skills, and mathematics are offered in this manual. An overview discusses the need for standards and includes a brief description of current practices in establishing them. In light of related social and political issues, it is advised…