I love exams, they are the best way to show just how much material you are able to remember, learned by rote and only needed for the shortest amount of time.
You see it all the time, most often right before exams where those students are huddled around lecture notes, cramming in as much of the content as they can before walking into the exam room, to recite it onto the paper. Most of it to be forgotten as soon as they leave that room 2 or 3 hours later.

I have come to find that this information is generally stuff that while important in terms of a specific field or situation, one does not need to carry with them for extended periods of time. eg Inserting and deleting nodes in linear and exponential hash tables or B+Trees.

Case in point, if i’m writing that new funky back-end storage structure for my latest coolMcCoolCool app, im likely to refer to a nice big fat reference book in-order to implement the data structure (one of my favourites atm is Algorithms in C by Robert Sedgewick ) or the internet in general as it also contains a whole wealth of info on great structures. What i try to do as a general rule of thumb, is to not assume that something that i learned in a uni subject is slightly relevant.

This has come about mainly because I’m fairly sure that anything that we learnt in a uni class, while fundamentally correct has been been simplified to such a level that it is in no way an optimal solution. Implementing it would be sure to highlight these limitations in quick step. Instead I’ll refer to the books and hopefully find something that will work well for the given situation and has an execution curve that is less than exponential as the number of objects indexed increases.

I guess i am just really going on about nothing, as exams have been around for atleast 100 or so years. Yet i have to ask myself. Do exams really work for modern subjects, in a world where information is readily available if you just know where to look?

Surely it would make more sense for the majority of a given subjects assessment to be done in terms of using research materials available to ones-self in-order to find solutions to problems independently. rather than sitting exams where we are expected to recite material learned in lectures.
What does having 60% or more of a subjects end score based on a single 2 or 3 hour exam paper have to do with anything other than making it easier for the subject organisers to assess the students enrolled in the subject. Any results given are based not on ones ability to find solutions to problems, but on their ability to remember some semi-relevent fact or method learned by rote that was mentioned in passing 10 weeks ago, in a single 2 hour lecture.

This is the sort of thing that I am finding all to often at the University level. Surely what one should take away from a 3 or 4 year degree, especially in rapidly changing areas such as Computer Science is more than the material taught in the subjects that make up the degree. It really should be about knowing how to find information for an unknown problem 5 years down the track, and the ability to apply ones-self to learn ways of solving problems that didn’t existed back when the degree was being undertaken.

Just a few thoughts that hopefully have an effect on someone in the near to long-term future.

-Rich

This entry was posted on Sunday, June 3rd, 2007 at 4:43 pm and is filed under General. You can follow any responses to this entry through the RSS 2.0 feed. You can skip to the end and leave a response. Pinging is currently not allowed.