Fandom at work

[shadowscast]

On an intro to stats test I gave yesterday, this was question #1:

The police department in Sunnydale wants to know how the adult population feels about female police officers. The head of public relations chooses 100 residential addresses at random from the phone book, and sends a uniformed male police officer to interview the people at those addresses. The officer conducts the survey between 9 a.m. and 5 p.m. on a weekday. He finds someone at home at 26 of the houses. The question he asks is: "Considering the recent rise in violent crime in Sunnydale, and considering that women on average are smaller and weaker than men, do you feel comfortable with the police department hiring more female officers?"

a) What is the population? [1 mark]
b) What is the sample? [1 mark]
c) List all the possible sources of bias in this scenario. [4 marks]


This is how I amuse myself! (And, hopefully, at least a few of my students.)

Comments

[yourlibrarian]

Well you've amused me :>

[shadowscast.livejournal.com]

Mission accomplished! :)

[shapinglight]

I don't understand!

:Wails:

I think it's just too mathematical for me.

[shadowscast.livejournal.com]

Aww!

Basically questions a and b are just technical little things (first someone has to tell you what a "population" and a "sample") are. Question c boils down to "what are the Sunnydale police doing wrong?" which as we all know already, is quite a lot!

[shapinglight]

I think something weird happens in my brain whenever anything even vaguely mathetmatical comes up. It just seems to stop functioning.

Am glad my kids aren't like that. They take after their dad.

[julia-here.livejournal.com]

"Considering the recent rise in violent crime in Sunnydale, and considering that women on average are smaller and weaker than men, do you feel comfortable with the police department hiring more female officers?"


Oh, great example of a push poll question, there.

Julia, having done actual door-to-door survey work in SoCal I could give so very many answers to C, starting with the "poeople home at that hour are divided between the very old, the very young, the non-english-speaking and the chemically addled"

[em-fish.livejournal.com]

I was just going to say; "Why does the phrasing of the question make me want to punch someone in the face?" ;)

[shadowscast.livejournal.com]

Those were, indeed, two of the answers to question c.

Those must have been some long days for you!

[anelith.livejournal.com]

I've never taken a stats course! But let me see if I can guess. I work for a survey research company (part time as a database application programmer) so I have a little exposure to the concepts.

a) I think the population might be the adults of Sunnydale

b) I think the sample might be the 100 households

c) Bias -- now there's a fun one to think about. First of all you're leaving out anyone who doesn't get listed in the phone book. Secondly you're leaving out people who don't stay home during the working day. Thirdly the phrasing of the question is absurdly biased. A fourth possibility of bias might be whether the random sample was truly random or not. Speaking as a database programmer, I know that the word random is often tossed around freely when it has a very specific meaning. I've probably left some possible sources of bias out.


I've worked for my boss for years, but my current part-time-work-from-home-telecommute incarnation of this job started about 2.5 years ago, working for him at his own company. The first database app that I built for him in this new situation, I used dummy data that was all drawn from BtVS/AtS characters. When I went down to New Jersey to demo it, there was only one person (out of about 8) in this tiny company who recognized the names. A fellow fan!

Did any of your students recognize the name Sunnydale?

You did as well as a lot of my students.

[shadowscast.livejournal.com]

a) I think the population might be the adults of Sunnydale

Correct!

b) I think the sample might be the 100 households

Almost! Actually it's the 26 people the officer actually gets to talk to.

c) Bias -- now there's a fun one to think about. First of all you're leaving out anyone who doesn't get listed in the phone book. Secondly you're leaving out people who don't stay home during the working day. Thirdly the phrasing of the question is absurdly biased. A fourth possibility of bias might be whether the random sample was truly random or not. Speaking as a database programmer, I know that the word random is often tossed around freely when it has a very specific meaning. I've probably left some possible sources of bias out.

Very good! The only thing you missed was that with a uniformed officer standing in front of them asking the question, some people might be intimidated into changing their answer. (As for the random thing, since the question said it was random you can assume it really was; I talked about proper random sampling techniques with the class earlier, with random number tables and such.)

So your total score on the question is 4 out of 6. :)

Did any of your students recognize the name Sunnydale?

A few weeks ago I wore my Puppet!Angel t-shirt while teaching, and one student came up to me and identified it. So I know he got it. Nobody else said anything!

Re: You did as well as a lot of my students.

[anelith.livejournal.com]

That was fun! :-)

[slaymesoftly.livejournal.com]

ROFL - what a fun teacher you are!

PS - have you checked my LJ lately?

[shadowscast.livejournal.com]

No, I hadn't! (I'm not so much "behind" on my flist as "completely gave up on trying to stay up-to-date, quite some time ago.") But now I have ... thanks so much! (Longer reply to follow in your LJ.)