Re: Take my survey?
- From: "Bo Raxo" <crimenewscenter@xxxxxxxxx>
- Date: Mon, 10 Mar 2008 14:39:38 -0700
"Kris Baker" <kris.baker@xxxxxxxxxxx> wrote in message
news:WQZAj.12444$Ru4.10527@xxxxxxxxxxxxxxxxxxxxxxxxxxxxx
<mylifeofcrime@xxxxxxxxx> wrote in message
news:2fcac5cb-6d2b-4fcc-9b48-d1b3ee4f7095@xxxxxxxxxxxxxxxxxxxxxxxxxxxxxxx
On Mar 9, 11:45 am, "Kris Baker" <kris.ba...@xxxxxxxxxxx> wrote:
"BrennaNE" <Brenna.Ehrl...@xxxxxxxxx> wrote in message
news:a56e4a44-637d-42f0-ad6f-47796490788a@xxxxxxxxxxxxxxxxxxxxxxxxxxxxxxx
Hey all!
I'm working on a project for class on crime blog posters. Would you
mind taking my survey and helping me out?
http://www.spammonkey.com/s
Yes, I would mind. If you are truly in "class", you need to
change schools. Find one that teaches statistical sampling
techniques, and the dangers of self-selecting samples.
spammonkey.com seems to be spamming a lot lately.
Kris,
She is for real. She goes to a very well known school as well. And
according to my advisor in school, there is nothing wrong with the way
she is doing her survey. It is actually very common now. Especially
since she did not say only a specific type of people here could do
it. It is open to all. If she went to her school and tried to find
out about crime readers/blog commentors, it may not work well at all.
That would be skewed as well, as maybe no one there reads about crime
and so her survey would show that no one reads crime blogs and that is
just unrealistic (especially with my blog stats). In this group,
there can be many that do read crime blogs and many that don't. Not
everyone interested in crime reads blogs either.
-----------------------------------
She may be for real, but she doesn't know this isn't a blog?
You really cannot base opinions on self-selecting surveys
that are not statistically-based., I spent years in statistical
occupations, and saw how skewed results could be when
you asked for people's opinions.
It's like all of those Ron Paul supporters who filled up all
the self-selecting polls with their perl-script votes; it
xkewed those results.
A true statistical sample could be by counting the regulars
here, detemining one's required statistical accuracy,
determining the correct statistical sample size, randomly
selecting the names, and then posting a note here saying
who'd been selected and asking for their help in this
assignment.
It'd be more educational to do it that way, even if the
other way is easier.
Easy wiki here:
http://en.wikipedia.org/wiki/Sampling_(statistics)
Kris
Just my background coming out ;)
She might have done a version of that. She sent an email to me about the
survey, and the email addresses of what I presume are regular posters around
here were in the address line. I wrote back, suggesting she should put
those in the bcc when sending something to a whole lot of people who don't
know each other.
My other advice was that she should tell people whether their responses are
anonymous (for example, does it record your IP address?). I would think an
issue like anonymity/privacy would be something any legitimate researcher is
required by their university's ethics code to reveal, at least when
soliciting research subjects outside the campus community (and probably even
there, too). But that's just speculation on my part, your school may vary.
Bo Raxo
.
- References:
- Take my survey?
- From: BrennaNE
- Re: Take my survey?
- From: Kris Baker
- Re: Take my survey?
- From: mylifeofcrime@xxxxxxxxx
- Re: Take my survey?
- From: Kris Baker
- Take my survey?
- Prev by Date: Re: Mom of Auburn Slaying Suspect Apologizes
- Next by Date: Teenage Murderers in Rural Texas
- Previous by thread: Re: Take my survey?
- Next by thread: police
- Index(es):
Relevant Pages
|
