Faculty Divided on Alternative GPA

Zoology Professor Karen Steudel regularly advises about 50 zoology and honors students in the College of Letters and Science. Her advisees, she says, would benefit from an adjusted grade point average as a supplement to the traditional GPA.

“They want to take certain competitive courses, but some are afraid to take them because they feel their GPA may suffer,” says Steudel, a UW–Madison faculty member for 20 years. “I feel that’s a choice students shouldn’t have to make.”

As a business and statistics professor and associate dean of undergraduate programs in the School of Business, Robert Miller advises the Business Honors program, which includes more than 50 students, and meets with about 25 students every week. He says an adjusted GPA would complicate the grading process, rather than help students.

“It would be an administrative nightmare, and I don’t think student behavior would change in terms of taking challenging courses, except that they would complain more about their grades,” says Miller, a 30-year faculty member.

The views of Steudel and Miller illustrate the divergent feelings among faculty about grading – and in particular about an adjusted GPA, also known as an achievement index, developed by a statistics professor at Duke University. The topic arose in September, after Duke professor Valen Johnson asked UW–Madison and 15-20 other universities to participate in a three-year study of adjusted GPAs.

Johnson says his achievement index, or AI for short, will benefit students who take difficult courses and reduce the incentive for faculty to inflate grades. He uses a complex algorithm to calculate adjusted GPAs by examining the courses and grades of all students during a particular semester and academic year. In AI studies conducted at Duke, students at the extremes of GPA and class rank changed very little, while students in the middle varied greatly. Johnson hopes to add the AI as a supplemental measure of student performance at Duke.

Johnson’s request to participate in his study was referred by the University Committee to the Undergraduate Education Committee, which discussed the proposal Oct. 21 and is expected to issue a recommendation at its Nov. 18 meeting, according to Miller, the committee’s chair. The recommendation will then go to the University Committee for review.

Steudel, a member of the Undergraduate Education Committee, and Miller outlined their views at the Oct. 21 meeting and in subsequent interviews.

Steudel, who supports participation in Johnson’s study, says the traditional GPA is an important indicator for success in future employment and graduate school, but it has its limitations. Employers, who are aware of grade inflation at most colleges and universities, would especially be interested in the AI, she adds.

“The AI is the one thing we all want to know: how students compare with other students,” says Steudel, who grades using a curve. “It seems to me that this is an idea whose time has come.”

Miller, however, is concerned about the expense of compiling multiple submissions of data for the Duke study and the cost of adopting the AI at UW–Madison, which has about five times as many students as Duke. But beyond cost, Miller worries about how Duke may use UW–Madison’s data and if it will truly be confidential. The AI itself seems rigid, he says, and he questions how it will account for students’ maturity during their academic careers. He predicts the AI, if implemented, could collapse under its own weight.

“As someone who works in the trenches, I know that grades are changed frequently, that incompletes are changed all the time,” says Miller, who grades using an absolute scale. “If you make a grade change for one student, conceivably you have to make changes for all students with the AI.”

As of the end of October, three universities had sent or are sending data for the study, Johnson says. Another university has expressed interest, but cannot participate until next year due to a major software upgrade on its campus, he says. He declined to identify which universities are participating.

Should UW–Madison decide to participate in the Duke study, adjusted GPAs would need to be updated each semester, Johnson says. No data submitted for the study or any results would be made public without the consent of top UW–Madison officials, he says.

“We have no interest in reporting specific details of grade distributions at Duke or other participating universities, and (we) are very aware of the need to maintain the confidentiality of the grading distributions of individual instructors, departments and universities,” he says.

Johnson says he is interested in determining if UW–Madison students would report the AI to prospective employers and graduate schools and if faculty view the AI as useful. Not all colleges and universities, he adds, would benefit from the AI.

“So one of the purposes of this study is to see if the achievement index is palatable as a supplemental measure of student performance at Duke and elsewhere,” he says.