Exam Results ~ Normal Distribution

Friday 2nd May 2008, 10:35 am

Note: I was in the middle of Biotechnology Departmental Meeting when I started writing a draft for this post.

I was supposed to type down the minute of the meeting this time (as I am still the departmental secretary). Yeah, I was focusing on the meeting, but now I don’t have the feel of doing it. The lecturers have started arguing each others on the matter of Final Year Projects grading. Guideline has clearly stated that there’s certain limitation in the number of students who should get A. The Dean himself has challenged any lecturers who would like to defend that his/her grades which comprises a high number of A students to bring all the students to him and let him grade them. Let the dean evaluates their reports and let them present in front of the dean. Tough, right?

Well, let’s have a break. I would only type down something which is well worth and most important, unanimously agreed by departmental members.

Normal curve

What does it mean? Does it mean that it's unethical for lecturers to normalize the graph, i.e., in case of too many students get A, by multiplying the grades by certain factor? Well, it can't be helped since most institution would not accept it as normal if there are too many students get A. For me, it's true that everything occur would likely to follow the normal distribution. And when certain thing does not follow the rule, it shouldn't be forced to be normal. It's a matter of how to make it naturally normal , not by force.

I would humbly say that, if the grades are not normally distributed, it's not students' mistake, but lecturers'. Lecturers should know how to set the question based on their students' capability, so that the results would appear in a normal distribution. It's true that very tough to have a situation where, let say, 50% of students gets A. If it happens, then there must be something wrong with the questions, i.e., perharps too easy, or rather the same questions of previous semester. Lecturers have the guideline on how to set the questions. For example, there must be few questions which could only be answered by very smart students, i.e., only 10%. The rest should be average, as well as easy questions. To force the graph to be normally distributed is not the correct way.

(11:05) Huh.. they were still arguing. When were they going to stop? Could they just agree on a mutual decision? Or was it just because some of them who still din't (or didn't try to) understand the issue and guidelines provided by the kulliyyah or rather didn't want to acknowledge their mistakes? At one moment, seemed that all of them were talking. Then how should I catch what they were talking about?

(11:20 am) Alhamdulillah, at last they managed to stop it. The HOD stressed that lecturers should thoroughly study thesis/report assigned to them, page by page, and consider them as an amanah from Allah before coming to presentation. Marks should only be given according to the overall performance of the students.

Summary: We (lecturers) have not been doing our job well because it’s still full of emotional elements attached in it. A rational suggestion: go back and read the guidelines, and consider this as a normal course and treat it fairly as an amanah from Allah.

----- and the meeting has adjourned at 1:00 pm (it began at 9:40 am)