The concept of RG-giri

Two guys are out hiking. All of a sudden, a bear starts chasing them. They climb a tree, but the bear starts climbing up the tree after them. The first guy gets his sneakers out of his knapsack and starts putting them on. The second guy says, "What are you doing? He says, "I figure when the bear gets close to us, we'll jump down and make a run for it." The second guy says, "Are you crazy? You can't out run a bear." The first guy says, "I don't have to out run the bear. . . I only have to out run you."


That is the story of relative grading in a b-school… which I have consistently failed to understand. Not that I am a hypocrite who condemns a system even though it has actually helped me escape sure fail grades in many a courses. But then the whole logic evades me.

When you select people into a particular course you do what is normally called a normal curve fitting. The idea is that the set of people who apply are so large that they can be considered to be a normal population. Say for example they apply to a college that they know would take only based on their marks and that the application fee is a significant sum. I don’t think a sensible guy who has perhaps made a low score would apply. But still the population who apply is so huge that they fit the normal curve… ?

Ok… let us assume that they do satisfy the normal curve condition, that is some stragglers do apply for that college and form the tail end of the spectrum. Now what is the selection process that is normally adopted ? the top 1% or at max the top 10% are taken in for the course. Now where would that fit in the normal curve ? ( it is the shaded area)



Well actually most of the colleges admit in the top 1% only and that normal curve would look like this.



For ease of understanding and clarity of vision let us use the top 10% curve.

Well, so now we have a group of people who are actually top in the merit list. Of course the bottom of the group is bloated because they are not actually the worst of performers. I don’t think I have to explain the normal curve here. Only thing that needs to be understood here is that even the last person falls into the 99 percentile category and is a very intelligent student.

But what sadly happens is that the grades are probably decided by non-statisticians or we just blindly follow the theory of company policy and we try to fit a normal curve for these 1% of highest performers in the country. Will it be a good fit let us see in the figure below…



As we can clearly see the gap of white area that exists that does not overlap the actual distribution of the students and we have to somehow fill in the blue paint into the white area… now are we doing justice to the low lying areas of the original curve by pulling them up and calling them average, or are we doing enough justice to the people in the bottom of the blue layer who have been unlucky enough to be left out in the tail end. Of course I agree that they scored low marks in a couple of tests, but then was the older blue normal curve wrong should it have been a series of normal curves – like a sinusoidal pattern ?

Anyways this is something that is probably not going to change in a while and since the affected parties are also quite happy with this kind of distribution I don’t think I can do anything about but crib whenever I make a bad grade and feel happy for myself and offer my sympathies when I find that I was “lucky” enough that there were some people who have been pushed to the tail end…


Well what is the solution ? one simple solution is to have absolute grades. this not only keeps the toppers in the top and the bottomers in the bottom, it also gives a clear measure of the student's understanding of the concepts in the course... ok well to some extent as some of you may argue that the student would just mug up stuff and vomit on the paper... but then isn't that what is happening ? what is worse currently is that you do all that hard work of cramming crap into your head and manage to score a mark of 95 out of 100 say... u mite still be behind a 100 people who seemingly had more space in their head and could cram faster and hence make a C grade...

disclaimer: i am sorry for using the crammer word for toppers, but it is only an expression and nothing personal against anyone in particular. i myself wud sit in the bottom end of both spectrums - it is lonely at the top and the bottom too...