Grade distribution shows lower grades in STEM, intro level classes

.

.
.

 

Sunday, November 13, 2016 - 9:01pm

Over the past few semesters, average GPA's in the college of LSA were lower for STEM classes than for classes in the humanities or social sciences and social sciences, and GPA became higher as class level increased.

Last week, The Michigan Daily released a new web tool that allows users to view distribution data for all LSA courses over the past three full-length semesters. The data, obtained through the Freedom of Information Act, enables users to view grade distributions from courses and compare multiple classes on a single graphic interface.

Students and University community members can view data graphs of grades for specific classes and compare the grade distribution of one class to another. However, according to Economics Prof. Dmitriy Stolyarov, it is hard to attribute departmental grading policies to any overarching policies since grading is almost entirely at the discretion of the individual professor.

“Generally, the grading methodology is up to the instructor — this is one of the principles underlying academic freedom,” Stolyarov said. “Accordingly, there may be as many grading methodologies as there are teachers. These data are no big secret — every syllabus has a detailed description of how grades are determined.”

Grading practices, because they are decided by professors, are best evaluated by review of historical syllabi. However, based off the FOIA data, there are some macro-level conclusions that can be made about the difficulty or typical grades in a specific class.

STEM classes compared with humanities and social science classes

Students taking classes in STEM fields tend to receive lower grades, on average, than those majoring in social science and humanities fields, according to the data obtained by the Daily, which shows the historic grade distribution for all LSA courses for Fall 2014-Fall 2015.

As shown by the data, the average grades given to students in math, physics, chemistry and biology classes equates to a 3.1578 GPA. These STEM departments are among the most popular math and science departments offered by LSA.

In these departments, according to a 2009 study on grade inflation at the University of Michigan by former University Ph.D. student Alexandra Achen and Paul Courant, a professor of public policy and economics, the average GPAs are, in part, a reflection on the number of prerequisite courses.

“Strikingly, all of the low-grading departments at Michigan have large required courses,” Achen and Courant wrote. “Math, physics, economics, biology, and chemistry are consistently among the lowest-grading departments, and each has courses that are required for courses of study necessary for the professional ambitions of large numbers of students.”

In comparison, the class data also showed the average grades earned by students taking economics, political science, psychology, English, communications and philosophy courses — all humanities and social sciences departments — were equal to a GPA of 3.3793. This range between STEM and humanities and social sciences averages of 0.2215 is the difference between a B and a B+ average.

However, in the study, Achen and Courant were hesitant to attribute this solely to the quantitative nature of STEM courses.

They also noted that students enrolled in English 125, a large introductory English composition course, averaged lower than other English level classes. This, they argued, is because of the types of students enrolled in this class — which, according to Achen and Courant, is because there is a large number of students enrolled who will not go on to become English majors.

Another possible factor for the apparent difficulty of STEM classes is the commonly used factors that determine grades in these classes.

Biology Prof. Trisha Wittkopp who taught the popular pre-med course Biology 171, Intro to Ecology and Evolution, said she used a formulaic grading method for students determined through points earned from exams, discussion section preparation and participation, quizzes and iClicker scores, and does not round up. 

For classes like Wittkopp’s, she said, grades are determined by performance on largely non-subjective examinations and are reflections on test taking acumen, rather than more qualitative assessments like argumentative essays she said are common to non-STEM classes.

Introductory level classes compared with upper level classes

The class data also suggests upper-level classes have higher grades in comparison with introductory level classes.  

The trend holds true across most departments and academic disciplines. Among 10 popular and academically diverse departments at the University, the average GPA of 100-level courses is 3.2613 and increases to a GPA of 3.4572 in 300 and 400-level courses. Additionally, the standard deviation — a statistical calculation of the extent of variation within a dataset — decreases consistently as the course level increases.

According to Achen and Courant's study, this can be partially attributed to the major selection process — students in upper-level courses tend to be majors or minors in that field, and therefore more apt in such areas of study.

“Second, upper-division classes are likely to have higher grades than lower-division classes, both because students have selected into the upper-division courses where their performance is likely to be stronger and because faculty want to support (and may even like) their student majors,” Achen and Courant state in the report.

However, Mika Lavaque-Manty, the director of undergraduate studies for the department of political science and associate professor of political science, warned that this may be an overly simplistic view of grade distributions.

“People have different approaches to grading; some faculty may make a difference in how they grade intro courses as opposed to more advanced courses, some may not, but I know of no general approach,” Lavaque-Manty said. “I personally think it’s a conceptual mistake to equate the number of As with rigor, especially if the number of As are determined by prior distribution decision.”

Lavaque-Manty, who was also a member of the former Provost’s Learning Analytics Task Force, said he thought his own research into grade distributions also shows an overall GPA increase throughout the past two decades, regardless of department.

“A couple of years ago, I analyzed all University grades from academic year 1995-1996 to 2011-2012,” he said. “In that 15-plus-year period, average course grades went from about 3.35 to 3.45. That’s obviously some increase over time. Is it significant? Problematic? I don’t know.”

Despite this, Lavaque-Manty, a long-time critic of using grading as a metric for student learning, noted that the University should deemphasize the importance of grades.

“But I do think that your data suggest we are dealing with multiple currencies we think of as one: grades mean different things,” he said. “That’s why I think they are a very imperfect measure of anything — course rigor, student learning, student comparison — and as an institution, we would do well to start thinking of better ways of representing student achievement.”

Daily Staff Reporter Brian Kuang contributed to this article.