So I just had an awful exam. This is the same course where I failed the midterm. I may have passed; I needed 56%. Read more to find out just why this course sucked.

As a professional teacher, I expect the professor to know just a little bit about teaching. Apparently I am wrong. There are several major categories of tactics people use to learn. One, they listen to someone lecture. Two, they read printed material. Three, they ask questions of the teacher. Four, they learn by doing, aka homework.

Hardly anyone is proficient at all four of these.

The immediate conclusion from that statement is that if you do not make all of the material available through all four learning channels, guess what? People not proficient in the major avenue of information conveyance are going to suffer.

It's hard making the information available in all for channels. That's why high school material moves so slow: they engage in quadruple-redundancy teaching because the stuff they teach -- you know, like shakespeare and the history of the ottoman empire -- are REALLY IMPORTANT in your life.

Once you get to university, it is expected that people can effectively engage the material in more than one form. That way, things can move a little bit faster, and thus more material (and more difficult material) can be covered. The professor can choose to offer material in only 3 streams -- most often, printed material, lectures, and office hours for student questions. It is expected that university students can handle this; if they can't, they fail. That's just the standard that is applied.

The real trouble starts when the professor drops down to just 2 streams of information. There are people (like me) who simply don't do well simply by attending lectures. I have trouble staying focussed. Others can't learn well from books. In the rare case that the prof doesn't answer questions (and yes there are those kinds of profs, though rare) another subset of students struggle.

Well, this class I just finished taking -- Distributed Computing -- was like that. Fully 25% of the exam was only covered in class. The material wasn't in the book at ALL. The only printed reference to the material on the exam was a set of "bonus slides" which, upon examination, was just a series of diagrams with NO accompanying explanation. It is purely impossible to ferret out what exactly is going on in the slides. Several other exam questions were "As covered in class...". If more than 5% of the exam is "from class only", I'm sorry, but you're a fucking pathetic teacher. I counted 25%.

Secondly, the assignments, if you are going to have them, should relate to the class material. Guess what? That's right, the assignments were totally irrelevant. The second exam had as much work to get the REQUIRED ROBUST LEXICAL PARSER for the config file working as there was to the rest of the assignment. Furthermore, they didn't mark them in any sort of beneficial fashion: on one assignment I got -3 for "modularity". I assume he meant lack of or bad modularity; I'd argue that point in any case. He didn't say what about it he didn't like. That one work and the negative number was all he wrote on the subject. Maybe he had something against my naming convention? Perhaps he doesn't like that I write in pure C, and didn't write classes. Or maybe in a past life I ran over his dog. I'm putting my money on the last one, because I also got -3 for not specifying the program arguments in my documentation. The program arguments which were unambiguously defined in the assignment description. The arguments which my program outputs on the command line if you give it bad parameters. Despite the fact they managed to test my assignment flawlessly, despite the fact that everyone had the same fucking user interface because it was specified to the letter in the assignment, they couldn't seem to figure out the correct arguments to my executables. One thing is for certain: just because your solution gets 100% on the functionality section of the exam is no reason for them to think its actually a "good" submission. Actually, on the converse, it means they'll probably scrutinize it more for cheating; after all, students who sleep in class, don't provide program arguments, and write code with poor modularity must have purchased the solution off of an ex-student. Right?

Lastly, in a final year computer science course, the ratio of "explain" to "show" questions should be fairly low. This isn't a "theory" course. It's a "application" course. Almost all 4th year CS courses should be, because a majority of students are not (or should not) continue into graduate school. I respect the fact that the prof decided to do grad school, but that doesn't mean that he should tailor his class for grad students. I took the course to learn how to program network-aware applications; that at least would have been useful. Instead I know all about how to synchronize clocks. Unless CS students are taking specific "research" subjects, they should be taking all "application" in their final years. The university has a responsibility (and a vested interest) in preparing students to get jobs in the industry. I'm sure my potential employers will be happy to hear I know all about the different types of data consistency in caching system. You know, they'll be so much more impressed with than than say my final graphics project, my final AI project, or my final OS project.

But no, there were a lot of explain questions. Explain the difference between this and that type of replication. Name 3 leasing heuristics, their goals and their problems. As discussed in class, what is the problem with RPC in lossy environments, and what 3 methods did we propose to mend it?

Oh yes, and by the way, only 1 of those answers can be found in the printed material. To anyone who has trouble taking notes in class, here is some lube: this exam is going to be long and hard.

yeah, my fall semester I took a class called "programming at the hardware/software interface" but we spent the entire class learning C (which I mostly knew) and MIPS asm and all the programs were STRING PARSERS. even the final program we wrote in asm was a string parser (it was a regexp handler)

Posted by Cyb from 208.58.89.233 @ 08/02/2003 11:31 AM CST