[Balug-admin] evaluation data, etc. (2005-05-17 meeting/presentation, etc.)

[Michael Paoli]

Here's the numbers, ... or, ... lies, damn lies, and statistics ...
depending upon one's perspective ;-)

First, this is the text that was used for the form - printed two to a
letter sized sheet, and cut in half:

Evaluation Form: BALUG: LVM on LINUX by Michael Paoli 2005-05-17

Using a scale of 0 (worst) to 10 (best) how would you rate the
following:

Meeting time and location? ___ (able to get to it, find it, etc.)

Food and service? ___ (price/value, quality, service, etc.)

This presentation overall? ___

presentation materials? ___

oral presentation? ___

presenter's knowledge of subject/material? ___

examples/demonstration presented? ___

What do you think were the worst things about this presentation?

What do you think were the best things about this presentation?

Please provide any general comments on how you think this meeting and/or
presentation could be improved or should be changed:

(Optional) Your name and/or contact information:


Certainly feel free to reuse 'em ... just remember to change the
date/title line (or leave it blank and fill it in or update it as/when
appropriate).  Hey, maybe at some point BALUG will be so popular and
overrun with potential presenters, we'll have to thin 'em out ... and
objective data can aid in comparisons (also helpful for better adjusting
one's guesses of what the general and specific impressions were).

Anyway, first I'll give the items letter labels, so I can have a short
table reference for them (think mnemonics; if the 1st letter isn't
unique, think the 2nd ... follow the bouncing *'s ...):

L Meeting time and *l*ocation? (able to get to it, find it, etc.)
F *F*ood and service? (price/value, quality, service, etc.)
V This presentation o*v*erall?
M presentation *m*aterials?
R o*r*al presentation?
K presenter's *k*nowledge of subject/material?
E *e*xamples/demonstration presented?

And then the numeric data:
  avg. med. low hi. raws
L 8.7  9    5   10  10,8,8,10,10,5,9,8,10,9,9,,,
F 7.5  8    5   10  8,8,7,5,10,5,7,7,10,8,8,,,
V 8.7  9    6   10  8,,8,10,6,10,8,,10,9,9,,,
M 8.2  9    3   10  7,9,8,10,3,10,7,9,10,9,8,,,
R 8.2  9    3   10  9,7,8,10,3,10,7,,10,9,9,,,
K 9    9.5  7   10  9,8,7,10,10,10,9,,10,10,7,,,
E 8.4  9    3   10  8,8,10,10,3,10,7,,10,9,9,,,

Notes, etc. on the data ... any mathematician/statistician that spots an
error, feel free to correct :-) ...
Some of the forms had empty or non-numeric values for the numeric items,
those are treated as nulls for the sample data (aren't used in
calculating the average, etc.).
Numeric data on the sheets and (further below) text/comments written in
are my most reasonable attempts at interpreting the writing - there may
be at least some ambiguity in the written data (including the numerics).
avg. is arithmetic mean, a.k.a. average (add 'em up, divide by the
number of items)
med. is median - roughly speaking, half are above, and half below (sort
the values, if odd number of items, take the middle one, if even, take
the two next to the middle, and average them) ... anyone, feel free to
correct me if I don't have that done/stated quite right.  (median is
often relatively handy for seeing where most of the values are likely
clustered and/or offsetting small numbers of data points that are far
from the average.  E.g., if I have it right, 1,000 people, 999 of which
have a $10,000 income, and one of which has a $1,000,000,000 income, have
an average income of $1,009,990, and a median income of $10,000 ...
this is why, for example, one commonly hears reports on median home
prices, rather than average home prices).
low and hi. are the lowest and highest values anyone gave.
raws are the raw numeric data
I did check over the raw data fairly carefully (at least to the extent I
could best interpret the written digits), but I haven't double or triple
checked all the other calculations (although I tried to be fairly
careful, and hopefully didn't introduce error(s)).  The reported results
are rounded to the nearest 0.1 (all the numeric responses that were
provided on the sheets gave whole numbers (non-negative integers) within
the requested range).

And, the non-numeric feedback provided.  I did, at least
semi-intentionally, leave enough space around the numeric items so folks
could write some bits of comments in there too, if they felt so inclined
(but didn't explicitly ask for such there) ... anyway, some folks did
write some non-numeric stuff in there (generally along with numeric feedback).

Anyway, here's the non-numeric stuff that was written in, first with the
text they were (or appeared to be) alongside, and then what was written
in, ... less the (optional) contact information ... I also tried to
reproduce the written in text as literally as I could read/interpret it,
and where I felt compelled to squeeze in a clarifying comment, I put it
within angle brackets (<>):


Meeting time and location? ___ (able to get to it, find it, etc.)

Food and service? ___ (price/value, quality, service, etc.)

This presentation overall? ___

presentation materials? ___

oral presentation? ___

presenter's knowledge of subject/material? ___
does Know
I don't know enough to know!

examples/demonstration presented? ___
Make members learn you in control <this may possibly have been intended
to be associated with the item below, but the writing looks mostly
alongside this item>

What do you think were the worst things about this presentation?
lack of Depth
One thing: Michael talking to quickly + softly
Pacing and organization
No microphone
A BIT DRY.
No Mike will change!
Mike needs a Mike so he won't shout himself hoarse.
It was a bit over my knowledge
Above me -
Some digressions caused by members of the audience went on too long

What do you think were the best things about this presentation?
lVM on 1 Hd
giving Knowledge of LVN + its brethren
Content + knowledge
PRojEctoRs, Demo
INFORMATIVE.
doesn't read from display
LVM is very useful topic -
I learned alot of new things
Above me <literally as ditto (") marks from comment on the item directly
above it>

Please provide any general comments on how you think this meeting and/or
presentation could be improved or should be changed:
Need a mike
Microphone
rehearse it first
<starting here ...>
Repeat questions from Audence
Shorten Talk & Demo To less Then 75 min.
<... and through here, was written on back, so precise item association
is a bit ambiguous>
You could have used screen to have all your terminals in one window


Anyway, ... I'll comment a (brief?) bit on some of the non-numeric
feedback/comments ...
I think the (lack of) microphone comments pretty much speak for
themselves ... hopefully we'll have that covered for next presenter (at
least have it available if needed/appropriate).
... less than 75 minutes, ... talk + demo was actually significantly
under 75 minutes.  I think it was only a moderate bit over an hour, ...
talk, etc., started at approximately 8:05 P.M., and I believe Dick
called "time" at approximately 9:05 or 9:10 P.M.  Seems we might be able
to have a bit more time available for presentation, question and answer,
etc., if we can tighten up some of the meeting logistics scheduling.
Most notably, I note food doesn't seem to get served until about 7:30
P.M.  It's then a bit after 8:00 P.M. (about start of food arrival + 35
minutes) when folks are winding down sufficiently on their eating, that
seems presentation can start fairly well.  (my observation/theory that
folks pretty much get to visit/talk/chat from arrival up to presentation
time (and after presentation) - folks manage to still talk pretty well
through dinner ... so, ...
if food arrived earlier, I'd think folks would still get adequate time
to visit/talk/chat and such, and then there could be more time for
presentations and such).  Anyway, ... perhaps if we're able to get folks
to show up a hair more on time (we seem to be doing pretty good on that
presently), but more notably, get the food coming sooner (e.g. be able
to better estimate probable attendance earlier, so the restaurant can
get the first plates of food delivered sooner) ... anyway, maybe there's
some way of tweaking how that works :-).  Also, if food starts arriving
earlier, that might also help work as an incentive for folks to arrive
in a bit more timely manner too (particularly those more towards the
straggler end of the time range).  Anyway, ... just some of my thoughts
on tweaking timing a bit.
screen(1) - I didn't seriously consider screen(1) for several technical
and practical reasons, e.g.:
I needed to switch among GUI and non-GUI (e.g. xterm) windows,
I wanted to be able to go quite directly to specific contexts (e.g.
superuser (root) vs. non-superuser vs. various GUI windows),
with X (and window manager I was using) I could title windows to
distinguish them (e.g. "demo" vs. "demo#") hence selecting from menu the
window title and raising and bringing focus to that window in a rather
quick and efficient operation (e.g. Alt-F4, mouse movement or arrow
keys, <RETURN> or mouse click, and I was to any window, and it would be
raised and in focus - allowing me to switch among arbitrary X windows,
without additional steps to switch among windows within screen(1)),
screen(1) isn't useful for switching among sessions with different sized
(geometry) screens (varying numbers of lines and columns) ... though what
I mostly used for demo/display happened to have matched lines and
columns, my more densely packed notes for guiding me through the
presentation had significantly more lines and columns (reducing the
movements and additional steps I needed to do in that window - whereas
the presentation window's were optimized for projection readability).
Otherwise screen(1) is a great tool, ... and I do use it a lot (heck,
have two screen(1) sessions open, each with multiple windows within
them, at home as I'm typing this).

And, ... strange things one can do with statistics, ... If we look at
averages, I beat the food/service cold ... as did everything, really,
... though the food/service still had a respectable 7.5 average.  On the
other hand, the time/location trumped almost everything - I beat it in
one category and tied it in another, and it beat me on the rest.  Of
course, for time/location, there's also the self-selecting sample set
phenomena (e.g. those who have problems with the time/location don't
show up to take the survey).

Well, ... enough commentary/analysis on evaluation data ... at least at
the moment and from me.