Using GPA in Assessment: Cautionary Notes

The issue of GPA use in assessment plans has occasioned a good deal of discussion in the Assessment Subcommittee; the committee in fall 1996 reached a consensus on the following statements about such use:

1. We can make a useful distinction between grades earned within a program (i.e., grades in major courses taken within the department) and grades earned outside the program (e.g., grades earned in required courses taught in other departments or grades earned in internships, practica, clinicals, practice teaching, and so forth, where external evaluator[s] determine the grade).
2. The latter--grades earned outside the program--may provide excellent assessment of student learning outcomes; particularly useful are grades associated with internships (etc.) undertaken after the bulk of the "in-house" coursework is done and especially when those grades clearly relate to a specific and clearly articulated objective.
3. The former--grades earned within the program--should be used with care: (a) they should be, like external grades, linked to specific and clearly articulated objectives; and (b) they should be but one part of a varied set of assessment procedures and criteria, providing the program with useful comparisons. Comparisons between grades earned in the program and other assessments, for example, might very well provide a department with information about the soundness of its grades and, thus, allow it to address the ever-present concern of grade inflation. A very general GPA criterion--e.g., "Seventy-five percent of the graduates from the x program will have earned a 3.0 GPA in major coursework"--is never useful because it has no particular diagnostic power (i.e., the fact of meeting such a criterion points to no particular strength in a program, and the fact of not meeting such a criterion points to no particular weakness).
4. The essential question in relation to all assessment procedures and criteria--including those involving grades--is this: Do the stated procedures and their associated criteria actually and adequately measure or demonstrate what the program has projected as an outcome? To ask the same question another way: Will the data generated by these procedures and criteria provide to a program useful information about its students' achievements in relation to a particular objective?