ma179acl@sdcc3.UUCP (04/23/87)
In article <603@plx.UUCP>, ed@plx.UUCP (Ed Chaban) writes: > > Why do professors assign the *SAME* problems every semester anay? > Because they're just like you and me: *LAZY*!!! Let's face it, if they don't have to, why bother? -------------------------------------------------------------------- # Dave (a.k.a. Slash) # STAY HARD!!! # "Who is John Galt ?"- A.R. # "When can we all become one race?"-M.C --------------------------------------------------------------------
neves@ai.WISC.EDU (David M. Neves) (04/25/87)
>> Why do professors assign the *SAME* problems every semester anay? >Because they're just like you and me: *LAZY*!!! Let's face it, >if they don't have to, why bother? Sigh... Do you know how long it takes to design a good assignment? 1. An assignment should test and train the student on the material. 2. It should not be too easy or too difficult (in terms of amount of code and number of hours to do). 3. It should be understandable to the students. (For example don't give an assignment that depends on understanding the American tax system if you have foreign students.) 4. It should be interesting (!) and hopefully relevant (!!). 5. A variety of other constraints like: don't give assignments that require class supplied input files if many of the students have their own micros, don't give assignments that can be found in books I sometimes spend a couple of months thinking about an assignment. After I have the idea I then have to write down all the information that students may need. On brand new assignments I often find out later that I have left out 2 or 3 crucial pieces of information. Students really hate it when you change or update assignments after they have been handed out. I then have to figure out how to make it easy enough to do in 20-40 (or ?) hours time. When students enjoy an assignment and it gives them good training why through it out? I try to change it just enough to discourage (or detect) cheating. One problem with totally changing an assignment every semester is that the assignment may be too difficult some semesters and too easy others. For example, our database course here has 4 assignments and the 2nd one this semester took students over 200 hours to complete. If the same assignment is kept it could then be made easier. Because of the feedback given this semester the program description could be made better. I think that instructors and students both gain when assignments are improved over time. If you think making assignments is so easy why don't you you generate a complete new set of assignments for the class that you just finished? Make sure they fulfill all the constraints above. Good luck. -- David Neves, Computer Sciences Department, University of Wisconsin-Madison Usenet: {allegra,heurikon,ihnp4,seismo}!uwvax!neves Arpanet: neves@cs.wisc.edu
elg@killer.UUCP (Eric Green) (04/30/87)
in article <509@ai.WISC.EDU>, neves@ai.WISC.EDU (David M. Neves) says: > >>> Why do professors assign the *SAME* problems every semester anay? >>Because they're just like you and me: *LAZY*!!! Let's face it, >>if they don't have to, why bother? > > Sigh... Do you know how long it takes to design a good assignment? > When students enjoy an assignment and it gives them good training why > through it out? I try to change it just enough to discourage (or > detect) cheating. One instructor here assigns the same assignment every year in her assembly language class -- implement a simulator for machine "A" (insert handout here, with specifications of machine), and run it with sample machine language program "B" (insert another handout here). And do it all in the assembly language of machine "C" (was a Pyramid 90x for three years, now it's an IBM 3090). The assignment has much merit as a "weeding-out" period. People who cannot design software certainly aren't going to write a sizable assembly language program. The way she deals with the problem of programs from last year is to change up addressing modes, instructions, registers, etc. of the simulated machine, every year (and changing the sample machine language program for that machine, of course). > If the same assignment is kept it could then be made easier. Because > of the feedback given this semester the program description could be > made better. > > I think that instructors and students both gain when assignments are > improved over time. Definitely. Each year, her machine design is better specified (when I took the course, <n> years ago, she had a bunch of ambiguities and ended up having to write a new specification of the indirect addressing modes). It's very hard to write a specification that is clear and concise the first time through.... once you write it, someone inevitably says "well, what about special case #n? How do you do xyz? does it add the index before or after the indirection?" and other such words of praise :-). I've ended up re-writing major parts of BT's documentation after a couple of beta-tester customers got ahold of it and started on that (groan....). -- Eric Green elg%usl.CSNET CS student, University of SW Louisiana {cbosgd,ihnp4}!killer!elg Apprentice Haquer, Bayou Telecommunications Snail Mail P.O. Box 92191 BBS phone #: 318-984-3854 300/1200 baud Lafayette, LA 70509 Clever quote goes here, but I'm out of room!