Last week, in my History of Cinema III class, I gave a midterm exam. The exam was designed to have an equal number of questions from each week of class, and that the questions would weigh lecture material more than reading material. The exam contained thirty-six questions, six on each of our six class sessions, and consisted of 55% objective questions and 45% subjective short essay questions. The greater weight to the objective questions was designed to test students on recall of lecture material, and the subject portion would be designed to test student’s comprehension of major themes in each class session and was graded relatively generously.
The results of the exam were not too surprising for a class of sixty three students who took the exam.
The exam was worth 200 points, and as you can see, the average score was 148.61, which is a 74.3%, a C average. This is consistent with a large lecture survey course. Someone in the class scored 193 points, which was the highest score in the class. And the average students finished in just less than 52 minutes.
However, when we look at the grade distribution, the scores look a little less impressive.
One usually expects there to be a lot of Bs and Cs, and we indeed have that. There are however not very many As and quite a lot of Ds and Fs. I would have liked to have seen more As and fewer Ds so these results are a little discouraging.
Now, when I wonder what might have contributed to the lower than expected scores. I looked at two other metrics: how long did students spend on the exam, and how much did students come to class.
In terms of how long did students take on the exam, we see that the time spent on the exam was indeed significant. Once we discount the first person who finished, who performed rather well, we see that performance slightly improves when students spend a little more time on the exam. But then after a while, we see that performance begins to drop. Presumably, those who finished later simply struggled and tried to answer as many questions as they could.
Finally, since we cannot use attendance to factor into students’ grades, I wanted to see whether student attendance factored into their performance on the exam. Remember, I designed the exam to test recall on lecture material, partly to reward those who come to class. But let’s see if there a relationship between performance and attendance.
The above chart shows a strong positive relationship between attendance and performance. The small number of students who came to every class didn’t do as well as the many more than simply missed one class. But nonetheless, we see a strong trend showing that more attendance results in a better exam performance.
One thing to note is that I make podcast recordings of the lecture available, but if they listen without attending, it’s unlikely that it helps their performance. One question I would have is whether the students who attend a lot also use the podcast to review, but it’s not possible for me to tell at this point.
The conclusion that I see here is that aside from the one or two students who just “get it,” taking one’s time on the exam somewhat helps your performance on test day, but coming to class significantly helps your performance overall.