The Irascible Professor SM
Irreverent Commentary on the State of Education in America Today

by Dr. Mark H. Shapiro

"Disparities in student grading have led to a general degradation of America's postsecondary educational system.  Inconsistent grading standards result in unjust evaluation of students and faculty, and discourage students from taking those courses that would be of greatest benefit to them."... ...Valen E. Johnson.

Commentary of the Day - June 23, 2003:  Grade Inflation by Valen E. Johnson - A Review.

Grade Inflation - A Crisis in College Education by Valen E. Johnson is, in the opinion of the Irascible Professor, an outstanding contribution to the debate about the effects of ever rising grades in colleges and universities across the United States.

Valen Johnson is a statistician by trade.  Currently, he is a Professor of Biostatistics at the University of Michigan; but, before that he was a Professor of Statistics and Decision Sciences at Duke University.  In contrast to the many commentators on the subject who have spoken from the perspective of educational policy or educational philosophy, Johnson approaches the subject through the lens of the statistician.  The result is both unique and important.

This is a book with both strengths and weaknesses.  Probably the major weakness is Johnson's failure to chronicle the history of grade inflation in any great detail.  The introductory chapter instead focuses on the current state of the debate on the issue and on introducing the DUET (Duke Undergraduates Evaluate Teaching) experiment that was conducted during the 1998-1999 academic year.  Evidently, Johnson feels that the facts of grade inflation are so overwhelming that no examination of the data that supports the existence of the phenomenon is necessary.  Fortunately, those readers who want to see the data that confirm that grade inflation is common in American higher education can find ample resources at Professor Stuart Rojstaczer's web site

The second weakness of the book also is its main strength.  Namely, Johnson is an expert at mining statistical data, and at interpreting the results.  Those readers who have little or no background in statistics initially will find parts of the book tough going.  However, by subjecting both the data on student evaluations of teaching (SET) and the data on grades, themselves, to thorough statistical analysis, Johnson is able to reach firm -- and sometimes surprising -- conclusions about the effects of student evaluations of teaching on grade inflation and the relationship between student ability and grade distributions.

Johnson's book addresses five persistent "myths" about grades and grade inflation:

1. Student grades do not bias student evaluations of teaching.

2. Student evaluations of teaching provide reliable measures of instructional effectiveness.

3. High course grades imply high levels of student achievement.

4. Student course selection decisions are unaffected by expected grading practices.

5. Grades assigned in unregulated academic environments have a consistent and objective meaning across classes, departments, and institutions.

Johnson devotes two full chapters to examining the relationship between grades and student evaluations of teaching.  From an analysis of the more than sixty previous studies on this issue and from the results of the DUET experiment, Johnson not only shows that a strong relationship exists between expected grades and SET's; but, also is able to determine the most likely reason for this relationship.

The four most common theories to explain the correlation between expected grades and SET ratings are the teacher effectiveness theory, the grade leniency theory, the grade attribution theory, and the intervening factors theory.  The teacher effectiveness theory posits that students taught by more effective teachers learn more and receive higher grades as a result.  The higher SET ratings for these teachers merely reflect that fact.  At the opposite extreme, the grade leniency theory suggests that students simply reward teachers who are easy graders with higher SET ratings.  The grade attribution theory is more subtle.  It says that a student who expects a high grade in a class generally feels that the high grade has been achieved through his or her own effort.  At the same time, a student who expects a low grade is more likely to attribute the low grade to poor performance by the teacher rather than by lack of effort or talent on his or her part.  Finally, the "intervening factors" theory posits that factors that are not directly measured by the SET forms (such as the student's prior courses or interests) cause the positive correlation between grades and SET ratings.

Johnson is able to show quite convincingly that both the teacher effectiveness theory and the grade leniency theory are wrong.  Students don't automatically reward teachers who are easy graders with uniformly high SET ratings, and the students of teachers who receive high SET ratings don't learn any more effectively than the students of teachers who receive average SET ratings.  Instead, the data indicate that the correlation between grades and SET ratings is due to grade attribution and to a smaller extent to intervening factors.  In other words, instructors who grade more stringently are likely to have more students give them lower SET ratings than the instructors who grade less stringently, because they feel that it is the instructor's fault that they are earning a lower grade.

One of the most important conclusions from Johnson's work is that SET ratings generally don't measure student learning.  Instead, they measure student satisfaction with the teacher.  The two are not always related.  A teacher who is well organized and enthusiastic may garner higher SET ratings than a colleague who is less well organized and less enthusiastic in the classroom.  But, the students of the lower rated instructor may learn as much or more as the students of the higher rated instructor because so much of what is learned in college courses comes from work on assignments that are done outside of class.  A teacher who may be a bore in the classroom may assign more interesting or more challenging work to be done outside class.

Another important result from Johnson's work is that grade inflation is by no means a uniform phenomenon.  Although grade inflation has affected nearly all segments of higher education from the most elite private colleges and universities to the "open-enrollment" public community colleges, it has not affected all disciplines equally.  Inflation has been most prevalent in humanities and the arts, less prevalent in social sciences, and almost nonexistent in the sciences.  The result has been a Gresham's Law effect.  Students, and that includes the more able students as well as the less able students, are more likely to take courses in disciplines that have less stringent grading patterns than in disciplines that have more stringent grading patterns.  And, the less able students gravitate to majors where the grading is easy.  According to Johnson, this helps to explain the limited correlation between high school grades and SAT scores and college grades.

Johnson also raises an issue that few have considered.  Namely, that the differential grading patterns between disciplines creates inequities for students.  Those students who take more courses from the disciplines with more stringent grading patters will achieve lower overall GPA's than students who take more courses from the easy grading disciplines.  As a result we find that  premedical students tend to major in departments like psychology rather than in biology or chemistry because they know that they will be at a disadvantage when they apply for medical school if their GPA's are lower -- even though a biology or chemistry major might provide better preparation for medical school.

Finally, Johnson makes a number of suggestions for reducing the effects of grade inequity and grade inflation.  His most controversial suggestion, and the one least likely to be adopted in today's academic climate, is to weight student grades by a factor that takes into account the average grade in each of the courses that the student takes.  Thus, a student who takes mostly courses that are graded stringently would have his or her "effective" GPA raised; while a student who takes mostly courses that are graded easily would see his or her effective GPA lowered.

Grade Inflation is an important book.  Johnson does an excellent job of making the sophisticated statistical results accessible and understandable.  It should be read by every faculty member who serves on a personnel committee, as well as by all academic administrators.
Grade Inflation A Crisis in College Education (ISBN 0-387-00125-5) is published by Springer-Verlag (New York), and is available from your favorite book seller for about $25 plus tax.


Return to main commentary.

© 2003 Dr. Mark H. Shapiro - All rights reserved.