Page 2 - Trials of Teaching

missing the <point></point> ...

I had one like that.

We're standing around in this professor's office one day while she's grading papers from her Discrete Math I class. She picks up this one paper and then all of a sudden goes digging back through the stack. She compares these two papers for a minute then she hands them to me. I look and I see every single thing on the page down to the font, layout, paper type and printer ink is exactly the same. The only difference. I mean the only difference was the name.

So she's looking at us and she's completely flabergasted. I mean people work together on stuff (read cheat) all the time but usually they at least try to make it look different. She ended up deciding that since each of them did half the work, each could take half the grade.

Never let the cheats wear you down.

If collaboration is okay (and it isn't necessarily a Bad Thang), make that clear.

Virtually identical submissions should probably be treated as cause for failure. It takes a minimal amount of effort to phrase things in one's own words, even if the conclusions were arrived at in concert, in a spirit of cooperative help.

Strictly personal opinion, of course.

yes.. my cs lecturer made it clear any plagiarism would result in fail.. not just that particular assignment but the entire year. Cheats would have to (attempt) to re-enroll the following year. He recounted one episode with a tearful girl -

"but but .. I would have passed even if I hadn't handed in the assignment at all!!"

yeah, that's right ... duh!

When I was an undergraduate, in the '70s, a lot of my friends went to the University of Virginia. They had a clear and simple honor code. I'll paraphrase is as "A gentleman does not cheat, lie, or steal, or tolerate anyone who does"

This translated into a one strike and you are thrown out of the University, forever. They had serious Honor Court trials, due process, etc. but if you were found guilty of an honor code violation, you were thrown off campus, never to return.

It was a wonderful place. You could leave valuables, such as books, bicycles, cameras while you went off to play frisbee, and they would all be there.

I've heard rumours that they have weakened it in recent years, but don't know this for sure.

See that's beautiful. When I was in college there was always some teacher or administrator who would take pity on a student and the absolute most that would happen for anything would be they'd fail the class and take it again the next semester.

Those gutless professors and administrators are creating a moral hazard. By tolerating cheating with some light sentence, they are teaching that cheating, lying and stealing is acceptable.

You hear the term these days about letting the banks skip free from the horrible loans that they gave out -- let them free, and they will know that the government will bail them out when they make bad decisions.

The UVa approach was simple. Clear, unambiguous. Cheat and you are gone. Period. As a result, a few folks got thrown out of the University each year, and 99% of the folks did not cheat.

The statistics are that at least 40% of any university's incoming freshman class will not graduate. Its not clear that the cheaters would be the folks you want to hire or work with, so its hard to see that the UVa approach is somehow 'unfair'.

Thats because repeating a class means you're more than likely put back a semester (or quarter), so thats another chunk of tuition money.

I've been a TA for the past 3 years at the college level. I've only taught math classes, but one term I was the TA for a class on numerical analysis of differential equations, which of course involved a lot of programming. I didn't do a lot of teaching, exactly, but I did have office hours in which students came in with questions and buggy code (usually on paper). :-)

Students were allowed to use any programming language they wanted. Matlab and its various knockoffs were by far the most popular (naturally), but I also saw:

-Maple
-Mathematica
-C
-C++
-Python (with numpy and scipy)
-PHP

Yes, PHP. For numerical analysis. I only got PHP scripts from one student, though, and they were usually actually pretty well structured and readable. I was a bit surprised (but not too unhappy) that nobody gave me Fortran programs.

Unfortunately, there was no electronic submission system, so I never got to run students' code to see if it really worked except when they (very occasionally) sent me their buggy code via email to debug. Grading consisted of checking the submitted output to see if it was plausible and reading the code to find bugs. Thankfully, in most cases, it was pretty hard to produce correct output from a buggy program, so it wasn't too hard to separate the correct programs from the buggy ones.

I thought the code we get here at Dev Shed was ugly...until that class. I got everything from C programs consisting of a single flat unindented main() function with a few macros thrown in, to Matlab scripts with expressions spanning four or more lines with no white space whatsoever. On occasions like that, if the code wasn't producing the right output, I just circled the offending expression and wrote a note to the effect of "I'm not debugging this; find the problem yourself."

To be fair, I did get other programs that were clean and well thought-out.

The worst and most annoying situation was when a student would submit code along with output that I would eventually realize could not have been generated by the program submitted. That was the worst because I would pore through the entire program searching for the bug before realizing that there might not even be one!