Presenter:
Brian O’Doherty, East Carolina University
Description:
This session is about how to use educational and psychological theory to help managerial accounting professors better teach their courses and examine their students. The key ideas will be addressed through a discussion of how to build and use machine, or objective, tests. Much of the discussion will focus on the issue of test reliability. Reliability, in psychometric terms, essentially looks at the internal consistency of the test. Emphasis here will be placed on understanding the Kuder–Richardson 20 and 21 formulas. This will be followed by a discussion of the technical factors affecting reliability: quality of items; test Length; variability, and guessing.
The next issue addressed will be how to improve objective test items using the output generated by most test-scoring routines. The focus will be upon item analysis and test bias and how to avoid or at least ameliorate test bias. This will be followed by a discussion of assigning grades. Several methods of assigning grades will be discussed: reference-to-perfection methods, criterion-referenced evaluation methods, and norm-referenced evaluation methods. Some recommendations will be made about assigning grades and how best to combine inputs into a single grade. My examples will be from managerial accounting.
