Straighten Out the MSB Curve
Published: Thursday, September 27, 2012
Updated: Thursday, September 27, 2012 23:09
If the McDonough School of Business’ aim is to educate students to be among the most competitive in the nation and more attractive to potential employers, its grading policy does a poor job of furthering that goal. While MSB students may be appealing to prospective hires once they graduate, the grading curve creates severe problems within the school itself.
Some MSB disciplines, such as accounting, involve objective grading measures (i.e. are the numbers you gave on the test correct), while other courses, like marketing, are more subjective. Despite these noteworthy differences, all disciplines in the MSB employ the same curve: a department-mandated average GPA combined with a quota on different letter grades.
This would not be a problem in most classes based on objective grading, which makes every student easily comparable. But what happens when assigning numerical grades in a course is nearly impossible? In such cases, even small variations can have massive impacts on students’ final grades. The difference between a B and a B+ on a single assignment, for example, can be huge. And when professors have a large number of students, the impact that one grading decision could have on the entire class’ curve is not fully appreciated.
Yet this objective versus subjective grading issue is only a symptom. The root cause of most problems with the MSB curve is the large variability and disparity in the course difficulty. In notoriously difficult classes, a grade in the low 60 percent range may constitute a B, or even higher. As far as the MSB grading policy is concerned, this presents no problem: the distribution of grades will likely work out in such a way that the average GPA and quota requirements can be met. When classes are challenging, grades will more closely resemble a normal distribution because students have a greater opportunity to excel or to fail. Some students will perform much better than average and others will struggle, but most students will fall somewhere close to the median score, which is the most important measure when using a relative grading policy.
But what happens when a class is too easy? If the average in a class is 93 percent, the distribution of grades will not fall neatly into a bell curve. In these cases, outstanding students are no longer given the opportunity to stand out. In fact, two students with statistically insignificant differences in their scores could see very significant differences in their final grades. So while an A and a B+ may only differ by 3 or 4 percent, the effect of that difference on a student’s GPA is disproportionate to his actual class performance. Does a student who got one question partially wrong on a test really deserve almost a full grade point lower than someone who got the question correct? Objectively, their performances were nearly identical, but the results of the grading distribution say otherwise. If you accept the premise that grading systems ought to reflect student performance as accurately as possible, then there is no way that the MSB’s policy is ideal. If the goal of a grading system is anything but that, then institutional priorities need to be reevaluated.
So what’s my solution? It’s incredibly simple and feasible: the MSB should look at its historical grading data since the curve was implemented, find classes with unreasonably skewed distributions near the top and urge its professors to make those classes more difficult. While some students would bemoan the additional work, most would be happy with a system that accurately reflects their academic performance. An increase in difficulty could always hurt some students, but these students likely have artificially inflated grades in the current system anyway. For many students, fair and accurate representation of their class performance could mean the difference between getting their dream job and having to look elsewhere.
ERIC ISDANER is a junior in the McDonough School of Business.