exam 3 available for looking at

[Cara Stockham Kenny]

Hi everyone,

You can know see your exam at Gem Navier's office in ACES, on the 3rd
floor, across from Dr. McKinley's office.

Cara

[Adrian Kubala]

From looking at my test, I'd guess that the low average is mostly the
result of the strict grading of the short answer section. It's hard to
say how the questions were graded since there weren't any comments, but
it appears the graders were looking for lots of supporting details. Ok,
that seems expected, but for example: to answer the question about
whether stateful or stateless DFS is better in a LAN, explaining that
"stateful requires less traffic while stateless is more fault-tolerant,
so stateless is better in the LAN because LANs are faster but more
mission-critical, blah blah blah" was only worth 1 point; the other 4
came from explaining *why* stateful uses less traffic, etc. (despite
being irrelevant to the question of *which* to choose -- which the
answer from the answer key didn't even answer). If that's the answer
they wanted, why not just write the question: "Which of stateful and
stateless is more fault-tolerant and why? Which of stateful and
stateless uses less traffic and why?" Well, this seemed to be a general
theme, this is just the question I remember best.

I was also found surprising the decision to compensate for a difficult
Q5 (or was it Q4? whichever) by reallocating points, since this
penalized people like me who wasted a bunch of time on Q5 and actually
did OK on it. Alternatives which make me happier: making some of Q5
into extra credit, or simplest and least controversial, just appling a
curve to the overall grade. Though maybe that will happen anyways with
83 being the highest.