Recommendations for assigning weights to component tests to derive an overall course grade

Abstract

Background: Performance-based assessments are often not reliable enough to be used as the sole method for determining a grade. Purpose: Using more than 1 assessment format has several advantages and can improve the quality of the information used to calculate course grades. To achieve a valid and reliable total score representing a combination of multiple assessment formats requires appropriate weighting procedures. Methods: The classical test theory rationale and methodology for appropriately weighting in-course assessments is presented. Two reliability-related equations are used to make decisions about weighting for a computerized performance assessment and a pattern recognition multiple-choice exam. Results: Using the equations presented, the outcomes of various weighting scenarios are graphically presented. This technique produced weights that allow the instructor to obtain acceptable reliability while retaining a substantial emphasis on performance assessment. Conclusion: The weights assigned to component tests used to derive a total score have important reliability and validity implications. Course instructors need to consider both empirical reliability and logical validity evidence in determining component weights. When used in conjunction with classical methods, objectively scored item formats can augment performance assessments and enhance overall validity and reliability. Copyright © 2004 by Lawrence Erlbaum Associates, Inc.