Who’s better, who’s best?

I’ve mentioned earlier that I assemble my final grades out of a number of assignments, all designed to test different areas, where a range of marks are possible. There are very few 0 or 1 situations in the assignments, with the possible exception of participation marks (and these never make up a big percentage of anything). Now the natural spread of marks across all of these assignments contributes to a slightly noisy final mark. As a general indicator of performance, the band into which your final mark falls is going to be pretty reliable. If you get 50-64 in one of my classes then you know enough to move on from the course. You have enough knowledge. If you get 45-49… well, then we offer you a chance to sit the exam again or, sometimes, do more assignment work, because we’re sitting in a slightly noisy zone and giving you another chance might get you over the edge.

But what is the real difference between a student who gets 50 and one who gets 60? 64 and 65? 98 and 99? Which of these leap out to you and say “Well, obviously, this student is better!” What do we even mean by better  in this case? In my opinion, if a course is designed correctly, and the students are properly prepared, then a pass mark is available for everyone if they do the work and apply themselves. The higher grades are only open to people who move on to synthesis, or can demonstrate deeper understanding, or can communicate their ideas more clearly, or… you get the idea. It’s often a sign of increased effort, to a large extent, than greater ability. (I’m sure we all have very smart students who are mildly bemused as they fail again because they just didn’t hand enough up or actually read a book.) In terms of general bracketing, a student who achieves a credit has probably done more than one who has a pass.

Okay, yes, the borders are tricky but with enough assessment work and opportunity, sometimes people just don’t make it over the line. If we have the ability to fail people, then we can apply some sort of higher level banding to higher level concepts. (There’s lots to say here and I’m trying to keep this short and on point, but I’ll try to return to this later.)

When it comes to numerical marks, though, inside the bands, that’s a bit shakier for me. Yes, there’s a difference between 100 and 85 (do we need a VHD?), but 99 and 98? No, they’re pretty much the same. It’s a shame, therefore, that we often rank students in our heads on these numbers rather than their grades. Or we average the grades and come up with a number that is as accurate and as useful as an average can be. Should we be using the average? Mode? Median? Harmonic mean? What makes sense in the face of what these students have done?

It’s a shame then that values like these, or averages of these, or GPAs, have so much weight in the community. It’s not as if most people dig down, or have the ability to dig down, into the underlying courses to see what these things actually mean. Perhaps we should be awarding marks for mastery of core materials, a separate mark for performance in projects and yet another mark for advanced or extension work, tying it into industry practices and terminology? Then you could look at the numbers, as they are, and having isolated the components that have been jammed together before, get an idea of what that number means.

Imagine that a student has a perfect GPA for a set of courses that, as it turns out, have nothing to do with what your company does – except that the name sounds about right. Another student with a slightly worse GPA has perfect marks for core, including the language and techniques you want, and is fantastic on projects. You may find this out at interview or in the application but, if you only take perfect GPAs and ditch everything else, you’ll never even see number 2.

I’m not saying that our current system is unworkable, because it obviously works, but I have started to wonder what else we could do – in an ideal world. There’s no solution here, this is not amazingly pragmatic thinking, which doesn’t take into account the fact that so many people who take our students want that number. But that doesn’t mean we can’t talk about it and think about it. Maybe there’s something better? The first step is to work out what it is.