Sunday, April 15, 2012

Thinking about thinking is very time consuming.  I am being consumed by writing exams at the moment.  Then I read an article about having computers mark writing essays.  What an idea.  I could sit back while the computer reads this pile for me.  Not only that, it would grade it too.  But then I continued to read....what a little bubble was burst even before I could get it in the air.....

But here’s another view: the fact that a machine can score your essays just as well as your human readers suggests that your human readers aren’t really doing a good job of reading the essays in the first place.  It suggests that having the essays you have on your test, and grading them in the way you do, is an utter waste of time, money, and effort.  The fact that you’re able to waste this time more cheaply by using a computer doesn’t transform it into a worthwhile activity.

So dear group, just to throw this out there, do we have to have real live people grade our essays?  Another thought was from a teacher who would have his students run their essays through a computer and he would only read those that scored a B or better.  He was using the computer for screening purposes only.  I am not sure where I am, I only know that I have a pile, and it isn't getting any smaller for all my talking.  


  1. That is really time consuming Adria. I just can't imagine that a machine or a software can check essays and grade it as well. I just got tired thinking about thinking if this could be a reality one day. Even if it did come to be true .. how will this grading and reading take place, do we have to give some criteria to the computer, are we going to input the rubrics, what if there is a statement or a paragraph that is written in wrong language, how the machine will correct this and grade it .. I got headache already !!! Nice thought Adria.

  2. This is a pretty cool topic which I've never been able to properly explore. Any task that can be represented as an algorithm can be done by a computing machine. In marking papers, there should be some sort of marking scheme or rubric which markers use to make scoring decisions and which should provide consistency across markers. That rubric is an algorithm which, in theory, could be used as the basis for machine grading.

    Now, of course, it is easier to do with some types of marking than with others (it's been done with multiple choice and short answer questions for a long time now). Essays are more difficult because the range of possible responses is very high and the rubrics less definite. But I think it could be done, if you have enough samples of essays which we already know the mark for. The thing you need is a neural network ( which you can train to do your marking.

    Good luck!