Survey Results Blurb by John

Finally.

First, I'd like to thank everyone who helped us, and everyone who
took the survey, on behalf of Judy Ghirardelli and myself.
Specifically, the crowd who listened to Judy and I argue over
questions of format in group mail, and most specifically Pam
Korda, Joe Shaw and Erica Sadun who helped us proofread both the
Survey itself and/or the results of the Survey just a few days
ago.  Their proofreading and advice on format was invaluable.

Also, a personal thanks to Judy for grading questions 092 and
093, which I flat out refused to grade, knowing what the answers
were going to look like, and realizing that _no_ computer routine
I came up with would ever handle the flood of information.
You'll know why when you look at them.  Also, for her help with
formatting the final results.

Secondly, an apology for having taken so long in producing the
results of the Survey.  I thought it would only take a week or
two at the outside, with the routines I put together for grading.
Also, an apology on behalf of Judy and myself for having created
the initial debacle with questions 051 and 066, with the
malformed multiple choices.  You wouldn't _believe_ how many
times that damned thing was proofread.  Luckily, all but one
person got back to me with corrected answers, so the damage was
minimal.

Thirdly, a partial dodge of responsibility for having taken so
long in grading the survey-- obviously, my cynicism was broken
when I came up with my estimates and wrote my routines.  Never
again. 

The routines I wrote worked almost perfectly for the yes-no and
multiple choice answers, probably because the methods for
screwing up the answers were limited.  Of course, those people
who (against my specific instructions) wrote out the words
"undecided" have my eternal gratitude fro making my life that
much more difficult.  And we're not even going to talk about the
impossibly large number of ways people found to mispell simple
names like Gawyn and Galad.  It was these mispellings that made
it necessary for me to actually grade most of the Short Answer
questions by hand, rather than by computer.  The computer made it
easier, but not by a helluva lot.

I'll say this-- in the unfortunate event that I'm still at
Bradley for longer than six months, this whole things stands a
good chance of being done entirely over the Web, because I'll
have the access needed to create a simple Web-accessed survey.
(Bradley is amazingly paranoid about the degree of Web access
they give people.)

Penultimately, the grading methods:
When in doubt, I used what worked best.  

Because of my vast familiarity with the vi editor and the
attendant ed-style commands, I used a _lot_ of brute strength vi
commands and ex scripts.  I also relied heavily on grep, and
resorted to perl a few times.  I think I started to lose it, just
a little bit, when I started using perl to generate ex scripts,
rather than just using perl directly.  (In the short run, it
saved time, because I knew _exactly_ what I wanted for an ex
script, but didn't want to spend the time to figure out how to
get perl to do the same thing.  I doubt it saved processor time,
though.)

Anyone who actually wants more details can ask directly.

Finally, the stats and the format:

We had, in total, 224 survey respondants.  Last year, we had
less than a hundred.  I shudder to imagine the responses we'll
get if/when we do this next year.

For almost all answers, the format is given as:
# votes		% of votes	answer

So you'll have three columns.  In all cases, the percentages were
calculated out of 224 responses.  However, for many questions,
not all 224 people responded, so the number of votes doesn't add
up to 224 for all questions, nor do the percentages add up to one
hundred.  C'est la vie.  I could have changed the percent ratios
for each question, but you'd have had to wait another week.  I
figured you wanted them ASAP, though.

Notable exceptions are the ranked questions.  Here, I constructed
a weighted average of each score, and used that.  Remember, lower
numbers for the rank questions are better.

Also, I bowed to external pressure, and listed all answers (aside
from yes/no) in order of descending popularity.  I did, however,
move answers of the "No clue" or "None of the above" variety to
the bottom of the pile, followed by answers that were in-jokes.
(Ie, answers relating to me, my non-existant minion, Bela, or
other group regulars.)  The _really_ stupid answers were subject
to the Central Authority Censorship.  Ie, the vanload of hippies
from one answer, as well as certain others, went directly into
the bit bucket to wail and gnash their bits.  Such is life.

And now, the results.
My commentary will follow later, when I get around to it.

-- 
John S. Novak, III       	jsn@cegt201.bradley.edu
http://cegt201.bradley.edu/~jsn/index.html 
Most Easily Annoyed TA on the Net.