When Engineering freshman John Poruthur stepped into his math recitation last week, he was enthusiastically waiting for his teaching assistant to return the week's problem set, certain that he would momentarily be receiving his first 10 on a homework assignment.
However, Poruthur was shocked to discover that he had only received a 7, even though he had worked collaboratively with a friend from another section who had already received a 10.
Although initially puzzled by the difference in grades, Poruthur ultimately concluded that the discrepancy was attributable solely to the fact that different TAs had graded essentially the same work from two different students.
"I handed in the exact same homework as my friend from another section," Poruthur said. "It's really unfair that my grade for the semester will probably be lower than his -- even if we keep handing in the same homework -- just because I have a harder TA."
Although the marked disparity between Poruthur's and his friend's grades may seem to be an anomaly, Poruthur's experience is not as isolated an incident as it may seem. As the midterm crunch sweeps across campus, students in all of the University's undergraduate schools have been finding themselves in similar situations, exposing the growing problem of inconsistent grading by TAs.
"There's definitely some variation in grading," Associate Director of Penn's Undergraduate Mathematics Program Edward Crotty said. "We try to combat the problem in a number of different ways, but unfortunately, grading variation is a natural occurrence."
One of the principal approaches that the University has adopted to combat the problem of variations in TA grading is the requirement of a three-day training program, which all College TAs must attend before the fall semester.
In addition, a number of other potential remedies are utilized to minimize grading discrepancies. These include group grading, curving by TA section and TA grading of specific exam questions across all sections.
"There are lots of different creative methods that people implement to reduce grading variability," said John Noakes, associate director of the Center for Teaching and Learning, which runs the TA training program. "But the [School of Arts and Sciences] training session really tries to make grading more fair and equitable across the board."
Another frequently employed approach to combat grading variation is the faculty's utilization of strict grading rubrics, which provide grading criteria to be applied by TAs for students from all sections, in an effort to minimize the effects of TA subjectivity on grades.
"Most professors give their TAs pretty specific grading guidelines, especially for midterms and finals," Crotty said. "The use of these guidelines significantly reduces the overall influence of variability over major assignments."
However, while making the grading of major assignment noticeably more standardized, this policy does little to bridge the large gaps in the grading of smaller assignments -- such as homework, quizzes and lab reports -- which are graded almost exclusively by TAs. Although any single assignment of this type will not have as large of an impact on final grades as exams will, they can frequently constitute as much as 25 percent or more of a student's final grade.
"They should have stricter grading rules and rubrics for not only midterms, but for things like homework also," Poruthur said. "Everyone always focuses on being fair for midterms, but then other things can really get overlooked."
Although it would seem that courses like math, where answers are relatively objective, should be less susceptible to grading variation, many students said they have observed grading discrepancies to some degree in all subject areas.
"Contrary to intuition, the issue of grading discrepancies between TAs is definitely a problem for all subjects -- even math," College freshman J.T. Stinson said. "Because there were so many kids in my math class last semester, TAs would pick only two questions out of the homework to correct to determine your grade, and some of the TAs definitely picked harder problems, so their sections' grades were significantly lower."
Although no single cause can be pinpointed to explain TA grading disparities, professors and students agree that differences in past experience and background play a key role.
"The TAs are all from really diverse backgrounds and have had really different educational experiences," Wharton freshman Cameron Smalls said. "I think this really has an impact on their grading and causes some to grade more harshly than others."
Professors also noted that individual TAs may weigh some factors more heavily than others.
"Based on prior teaching experience, some people just have different attitudes toward grading -- either weighing the applied end or the theoretical end of mathematics more heavily -- which can definitely produce some variation," Crotty said.
And while diversity in TAs' attitudes and styles can be valuable by providing students with an array of educational experiences, most students feel that the faculty and administration should do more to guard against the grading inconsistencies which this diversity may cause.
Both faculty and students agree that the most effective way to minimize grading disparities may be improved communication between professors and their TAs.
"For the most part, I don't think the TAs correspond with the professors that well," Smalls said. "To ensure fairness in grading, you really need to make sure that TAs and professors are on the same page."Comments powered by Disqus
Please note All comments are eligible for publication in The Daily Pennsylvanian.