Passing Scores and Test Lengths for Domain-Referenced Measures

Abstract

U.S. DEPARTMENT OF HEALTH. EDUCATION & WELFARE OFFICE OF EDUCATION THIS DOCUMENT HAS BEEN REPRODUCED EXACTLY AS RECEIVED FROM THE PERSON OR ORGANIZATION ORIGINATING IT. POINTS OF VIEW OR OPINIONS STATED DO NOT NECESSARILY REPRESENT OFFICIAL OFFICE OF EDUCATION POSITION OR POLICY. Procedures for establishing standards and determining the number of needed in criterion-referenced measures are reviewed. The discussion of setting a passing score is organized around five factors: performance of others, item content, educational consequences, psychological and financial costs, and measurement error. Classical test theory, binomial, and sequential models IC) for determining test length are considered. 0 An illustrative table relating test length, proficiency standard, and required accuracy is provided. c.6.) FILMED FROM BEST AVAILABLE COPY'. In recent years there has been much attention given to criterionreferenced measures which relate test performance to absolute standards rather than to the performance of others. Popham and Husek (1969) provide a neadable account of the differences between such measures and the more traditional norm-referenced tests. The purpose of this paper is to synthesize much of the literature on establishing standards and determining the number of needed in criteriOn-referenced measures. This paper is written from the following perspective. A domain (i.e., population) of dichotomously scorable test items is conceptualized. This population of need not actually exist. Mat is important, though, is that it is described well enough so that a relatively high degree of agreement can be reached about which kinds of are or are not members of the population. In practice, only a reasonably representative sample of is required. 2 The of a domain may be heterogeneous in content, form and difficulty. In practice, however, they should be measures of a limited number of skills and knowledges so that it makes sense to establish a single proficiency standard. Of interest is the proportion of such a student can pass. It is assumed that some educational decision, e.g., the nature of subsequent instruction for the student, is conditional upon whether or not he exceeds a proficiency standard when administered a sample of from the domain. Thus, attention is directed toward the individual examinee and his performance relative to the standard rather than toward producing indicators of group performance.