DataGeek III: Analysing exam results
besides my job as a financial analyst, I am also holding lectures at a university of applied sciences about… well, SAP mostly. The course begins with some basic ERP theory, but the majority of the time is dedicated to providing first-hand experience of SAP ECC 6.0 for the students. After my introduction to the GUI, the students are to work themselves through a case study provided by the SAP University Alliance.
At the end of the semesters, this class is also required to take a paper-based exam. It deals with the three categories ‘ERP theory’, ‘Case study comprehension’ and ‘SAP GUI knowledge’. It took the opportunity to use SAP Lumira for an analysis of the exam results, trying to find out how grades evolved and in which area I have to enhance my lectures.
Evolution or Devolution?
I used a storyboard which two graphs which basically show the same thing: my very first class in the fall semester of 2010 gained the highest results, with an all-time peak at the highest grade, 1.0; the 3D Column Chart clearly visualizes this. It also shows that the lowest grade, 5.0, was almost on a constant rise since then. So high in the end, that the columns of the better grades are not even visible anymore.
The Bar Chart on the top right is shown in percentages. I am drawing the following conclusions from this: it also shows the remarkably good results of the fall 2010 semester at the top, then showing semesters where the distribution is more even. Coming closer to the present, it is also remarkable that the grades are not that separated anymore, visible through less different colors.
In a nutshell, I can say that the average results of the students in my lecture deteriorated over the years.
Save the Date
My university offers two exam weeks, students can rather freely choose which exam that take in any of the two weeks, or even to postpone it to the next semester. The first offered exam is less than a month after the end of the lectures, the second date is around three weeks later.
The Bar Chart indicates that the top grade occurs a lot more often on the first date, while the fails (5.0) are significantly more prevalent on the second exam date.
It would have been interesting to see how the results of students look like which are not taking the exam in the same semester, but one or even more semesters later. Unfortunately, my source data does not account for this.
Ups and Downs?
One of my prime interests for this analysis was how to improve my lecture. I categorized the questions into three areas, expecting to gain an understanding which area was answered best by the students, and which was worst, meaning that I need to take action for the lectures to come.
Fortunately or unfortunately, the data shows that there is little variation between the three areas. The Line Chart however emphasizes the constant decline of grade over time as already seen in the first exhibit.
This means to me that I need to improve in all three areas to make the content of the case study, the theoretical background in the SAP GUI look and feel stick to the students’ brain until and beyond the exam.