Very nice Alan. I too came out as pretty average (although I suspect I'm
better looking than the rest of these middle aged farts!!!)
Couple of obvious builds (I know you know but it's really for the benefit for
all those who have been recently distracted by the UPS thread):
1. Where's the median. You go through the mean mode and min/max but no
median. As measures of central tendency go you're in deep trouble with the
big bird of statistics - worse still - no mention of skewness and kurtosis.
I'm sure some listers are terribly disappointed.
2. I put on my stat hat (the one with the propeller on top) and determined
that your sample reliability is about 7.5% (that is your response accuracy
provides a gaping plus or minus 7.5%) - By example, from the sample of listers
you got 80% married. In fact this means that somewhere between 71.5% and
87.5% of all listers are married. Just think, on any given day there could be
a 15% swing in the number of us who are married - or divorced. Perhaps we can
work something out where on some weekends we are married and on others we are
not!!!!!
3. Of course I computed #2 at the 95% level of confidence (e.g., 95 out of 100
times you do this survey the percentage of married will be between 71.5 and
87.5%) - You could use the 99% level of confidence thus increasing the
interval to plus or minus 11% which means that there could be a whopping 22%
swing in the percentage of us who are married at any given time and we would
still be statistically legal!!!!!!. Think of the possibilities - when your SO
is not interested in the car show "you gotta go to" or will have a tizzy if
you work on that car this weekend, you could trade it for a not married
weekend and statistically it's within spec of the central limit theorem.
4. BTW, I did not use the finite population correction in my calculations
because your sample was less than 50% of the population being measred - good
thing to because it can really narrow that reliability interval a couple of
points.
5. Hey, how about some segmentation for sub group analysis - let's say on age
- hey I'm sure Kai wouldn't mind getting a couple of years on certain
weekends so he could legally drive (well, depending on progress to date, maybe
push) that British beast around a bit. And when some of the old farts on the
list cruise the high schools with the top down I'm sure they'd like to lose a
couple of years.
Hey maybe in the next survey we can include weight, hair and IQ as well (a lot
of horse trading can be done there, eh lads!!).
Ain't statistics and surveys a blast!!!
Seriously, great job Alan.
Dr. Doug
3 1969 MGs +/- 1 or 2
-----Original Message-----
From: owner-mgs@autox.team.net On Behalf Of Alan Lott
Sent: Monday, August 11, 1997 1:50 PM
To: mgs@autox.team.net
Subject: EXTRA! Survey Question Results
OK everyone, here are the results. First, I'd like to say "thank you" for
the high level of participation. Of the 391 people on this list, 140
responded. That is 36%, and as data gathering goes, is considered to be
"excellent" for returns to a survey. Second, the questionaire was the
"second to last draft" and not the "final draft." My mistake. In the final
draft, I fixed the USA only question about residence and corrected a couple
of other problems. Even though this didn't alter the results, I'm sorry if
I offended anyone. Lastly, All results reported refer to the "respondents"
of the survey. Statisticly, the results should be an acurate representation
of the list as a whole. Hope you all enjoy and if anyone wants the raw data
or more info, just email me! Oh, if the percents seem to fall short, it is
because I did not report the very low percentages for some questions. Enjoy!
mean = average
mode = most frequent answer
min = minimum value answer
max = maximum value answer
1) Age
mean 41.8 mode 40 min 15 max 66
2) Gender
96.4% of persons on this list are male...come on girls, get with it!
3) Marital Status (single, engaged, married, divorced; resist the
temptation to stray from these four choices :-) )
I should have broadened the scope on this question, sorry.
13.9% single .7% engaged 80.3% married 5.1% divorced
4) How old were you when you first became interested in Little British Cars
(LBC)?
mean 17 mode 16 min 4 max 50
5) How old were you when you first purchased an LBC?
mean 24.8 mode 19 min 14 max 66
6) What year/make/model was your first LBC? (yy/make/model)
mean year was 1966
Most popular: MG (72.5%)/ MGB (28.7%)
Second most popular make was triumph
Second most popular model was midget
Overall, 13 makes and 32 models were represented.
7) Including those you now own, how many LBCs have you owned in your
lifetime? (don't include never functioning/never will function parts cars)
mean 6.6 mode 1 min 0 max 114
8) What year/make/model was your FAVORITE LBC? (yy/make/model)
mean year was 1965
Most popular: MG (75.6%) / MGB (28.3%)
Second most popular make was Austin Healy
Second most popular model was MGA
9) How many LBCs do you currently own?
mean 2.5 mode 1 min 0 max 33
10) How many miles annually do you put on your most frequently driven LBC?
mean 6368 mode 5000 min 200 max 32,500
11) Are you, or have you ever been, a professional mechanic?
10.7% said "yes" 89.3% said "no"
12) On a scale of 1-5, rate how much of your own work do you do on your
cars? (1= none, 5= all of it)
mean 4.1 mode 4 min 1 max 5
13) How many AVERAGE hours per week do you spend working on your LBC?
mean 5.1 mode 2 min 0 max 30
14) Do you use an incandescent or flourescent drop light?
This was the S.S. Minnow of survey questions...doomed from the start. I
was just trying to (unsucessfully) inject a little humor given the recent
"shop light thread." I did, however, tally the results. The ambiguous
wording was fixed in the "final draft."
48.4% use an incandescent drop light
21.4% use a flourescent drop light
12.9% use both
and...
12.1% use neither
15) Considering all aspects (customer service, price, etc) rate the top
three (in order; good to bad) establishments you buy LBC parts from.
This was another S.S. Minnow question. I tried to pack too much data into
one question.
I lumped all responses and am reporting the top three of the 16 or so
establishments.
Moss 26.6% Victoria British 16.7% Abingdon Spares 16%
16) Rate your automotive "Purity" on a scale of 1-5, with 1 being = every
bolt and screw must be original, to 5 = anything goes (including major
metal cutting), as long as it works.
mean 2.7 mode 2 min 1 max 5
17) What city and state do you live in? (please us standard 2 letter
abbreviation for state)
First place goes to California, followed by Texas, and then Pennsylvania
18) How long have you been a member of this list? (use months, please)
mean 14.9 mode 12 min 1 max 60
I used 60 months (per a couple of old listers)for the people who indicated
they had been on the list since it was started.
Thanks again for participating!
Regards,
Alan Lott
GTA Dept of RSE
Auburn University
'71 MGB (rolling restoration)
"Diplomacy is the ability to let other people have things your way."
|