How COVID changed student performances

Most professors are familiar with the normal distribution. For college-level courses, we were led to believe that if we did our jobs appropriately, a course populated with a generally random group of students would generate a set of final grades that were, roughly, normally distributed.

This normal distribution of grades was looked at a lot like the Law of Averages: half your scores will be below your average, and half of your scores will be above your average. Both were inevitable if things were done appropriately across the board.

Example: Random Course 101

For example, Random Course 101 has a professor who has done as normally expected of a college professor teaching their subject matter. Their course was filled with a random population of 100 students across all demographics: age, gender, ethnicity, etc. Barring any other unusual factors, the professor would be right to expect the final grades in their course to fall within the normal distribution curve, as seen below:

China

A's: The professor, semester after semester, can expect approximately 2 to 3 As in Random Course 101. B’s: The professor can expect approximately 13 – 14 Bs in Random Course 101. C’s: The professor can expect approximately 68 – 69 Cs in Random Course 101. D’s: The professor can expect approximately 13 – 14 Ds in Random Course 101. F’s: The professor can expect approximately 2 to 3 Fs in Random Couse 101.

The normally distributed set of grades in a course will have the same number of A’s as F’s, and this won’t account for very many students in the course. Similarly, the normally distributed set of grades in a course will have the same number of B’s as D’s, and that will account for a little more than a fourth of the students. The normally distributed course can be expected to yield more Cs than any other grade, accounting for almost 70% of the scores.

Skews of A’s

We know this normal distribution does not always play out of course. Sometimes, the distribution of final grades in a course are skewed towards one side of the chart or the other. For example, Random Course 102 might have a higher percentage of A’s than what a normal distribution would yield. Traditionally, a skewed distribution of A’s could have a few different factors involved:

The course is a relatively easy course that most students will have no trouble with, such as Badminton 101. For such courses, a distribution of final grades that are skewed towards the A’s is expected.
The professor has made their course too easy. For example, we would not expect 80% of students in Calculus 301 to have an A. The professor would be advised to evaluate their practices.
Cheating – the course could have had many students who were able to cheat on their coursework or the calculation of their final grade, thus resulting in the unusual amount of high scores. You can probably think of other reasons why a course might have a much higher percentage of A’s than the normal distribution.

Skews of F’s

There are also courses that yield a disproportionate number of F’s than the normal distribution. Random Course 103 might have 90% of students end with a final grade of F. Obviously, this would be a skewed distribution that bears examination, and like the A skew, it can be caused by a plethora of factors:

The course is an advanced level class that students must absolutely demonstrate mastery of, or they will not pass. An example of this are certain medical courses, and such courses often have high failure rates. Indeed, we do not want to be put in a place of needing lifesaving treatment from a student who has not truly mastered their craft.
The professor has made their course too difficult or has not properly taught the material. English 101, for example, is a course that most college students will take. Across the entire spectrum of students, we expect such a course to be fairly normally distributed. An English 101 course that has 90% Fs needs to examine what is happening.
Professor error – the professor has made a mathematical error in the calculation of grades and needs to redo them.

You can probably think of other reasons why a course might have a much higher percentage of F’s than the normal distribution.

The great change

When the great change happened will depend on who you ask. For me, it happened at an NBA game. The night of March 11th, 2020, fans at the Oklahoma City Thunder game were told to “Exit the arena...” right before tipoff. It was a truly frightening situation. You could hear people screaming in the background. The announcers assured everyone they would be ok, but they couldn’t tell them why they had to evacuate. That moment quickly ushered in the realization that things were different now.

Very quickly afterwards, schools were shuttered. Roads were empty. I remember walking down the middle of the street with my son at noon on a weekday in April. Everything was closed and there was no traffic. A lot of people weren’t leaving their homes. It reminded me of something from the zombie apocalypse television show ‘The Walking Dead’.

Jobs moved online. Home offices were redesigned. College students went back home. Elementary kids learned what Zoom was. Families started daily walks together. Values began to be reevaluated. What was important in our lives? Was it our daily grinding commute to work? Was it the pointless layers of workplace and educational bureaucracy that we had become completely subservient to?

Somewhere during this, the dynamic between college professors and students changed too, and a new distribution of student scores has emerged.

The new normal

An anonymous analysis of several college courses has yielded a large number of courses that are experiencing a distribution of final scores that looks like this:

China

This new distribution of student scores post-COVID as shown in Figure 2 shows a visually dramatic shift from the normal distribution of student scores that existed in Figure 1 pre-COVID. Now, courses are routinely filled with many final scores of F’s and A’s, with only a spattering of D’s, C’s, and B’s. In random courses with 30 random students, it is now normal for college professors to see the following breakdown of final student scores:

A's: Professors now expect approximately 12 – 13 A’s. B’s: Professors now expect approximately 1 – 2 B’s. C’s: Professors now expect approximately 1 – 2 C’s. D’s: Professors now expect approximately 1 – 2 D’s. F’s: Professors now expect approximately 12 – 13 F’s.

When shown this very pattern of final scores from their own courses, one college professor of Health and Physical Education matter-of-factly noted that the distribution now resembled a ‘dumbbell’ – as in the weights one lifts in a gym. Not meant to be offensive but rather a descriptive representation of The New Normal, the dumbbell curve appears to have supplanted the normal distribution of final scores in college courses across the spectrum.

Analysis of the dumbbell

The dumbbell curve is now being recognized by college professors across the United States as a predominant distribution of final student scores in college courses, but what has caused such a paradigm shift in student performance? Although the research in this area is in its infancy, there is much speculation as to why this new normal is occurring:

Student apathy is at an all-time high. Many students are logging in to online courses, getting on the roster, and never showing up again.
Success rates are now being tied to Federal funding at colleges across the country. Faced with mounting pressure to raise these rates, professors have done everything possible to make their courses easier than ever before. Therefore, any students who are engaged in their courses tend to make an A.
On the community college level, many schools have gone to tuition-free models. Meant to boost enrollment from all demographics, many students find that they have no extrinsic motivation to succeed in their courses.
You can probably think of many other reasons why college courses now have a new normal of student score distribution.

The gunslinger

COVID changed the world. It changed the world in ways that were obvious, in ways that were not so obvious, and it continues to show us that it is still changing the world. Since COVID emerged, the educational world has changed, perhaps forever, and it continues to cough and gasp its way along, much like an ancient machine attempting to start back up. Educational institutions are notorious bastions for the old ways, slow to change, and even slower to embrace innovation. Students have been transformed from a naturally social construct to anti-social loners basting in the glow of their screens. I don’t know if we will ever return to the normal distribution or how long the new normal will be here. I suspect there are more challenges ahead as we are forced to deal with the harsh realities of what these new student distributions really mean, and as we begin to realize the pitfalls of tying government monetary support to student performance. In the end, it may just be that post-COVID, the world has moved on.