Statistics Research Paper

For the discussion of math. Duh.

Moderators: gmalivuk, Moderators General, Prelates

Magritte's Apple
Posts: 9
Joined: Wed Jun 18, 2008 3:38 am UTC

Statistics Research Paper

Postby Magritte's Apple » Sun May 02, 2010 6:39 pm UTC

Hello, all!

I am conducting research into political and economic views for a minor statistics project.

I need as many participants as possible to take my survey! Please do so! It's only 16 questions long. :)

http://spreadsheets.google.com/viewform ... QkFSX3c6MQ

Thanks!

User avatar
Kurushimi
Posts: 841
Joined: Thu Oct 02, 2008 12:06 am UTC

Re: Statistics Research Paper

Postby Kurushimi » Mon May 03, 2010 9:12 am UTC

I've only taken one AP Stats course so correct me if I'm wrong but, doesn't this create a survey bias? Like a nonresponse bias and then there's the fact that people on xkcd aren't exactly representative of the general population.

User avatar
the tree
Posts: 801
Joined: Mon Apr 02, 2007 6:23 pm UTC
Location: Behind you

Re: Statistics Research Paper

Postby the tree » Mon May 03, 2010 10:32 am UTC

Kurushimi wrote:I've only taken one AP Stats course so correct me if I'm wrong but, doesn't this create a survey bias?
Not much more than any other survey. And there's no definite reason why specifically reading xkcd would introduce that bias, unlike if the questions were about how people use the Internet or their opinions on science education for instance.

However, the OP may wish to ask a mod to move this to the general board which is where I'd have posted it.

User avatar
Velifer
Posts: 1132
Joined: Wed Dec 26, 2007 4:05 pm UTC
Location: 40ºN, 83ºW

Re: Statistics Research Paper

Postby Velifer » Mon May 03, 2010 11:42 am UTC

Yeah, massive self-selection bias, complicated by no way to even speculate on what direction that bias goes. The population in the survey frame isn't particularly well-defined (anyone who sees his link) and the respondents will be those who choose to click through AND complete the survey. That's not gonna be anywhere near random, even within the sampling frame.

The good news is that the OP can write the research paper on why their method sucks. That's an easy 50 pages.
Time flies like an arrow, fruit flies have nothing to lose but their chains -Marx

User avatar
the tree
Posts: 801
Joined: Mon Apr 02, 2007 6:23 pm UTC
Location: Behind you

Re: Statistics Research Paper

Postby the tree » Mon May 03, 2010 11:53 am UTC

Velifer wrote:The population in the survey frame isn't particularly well-defined (anyone who sees his link) and the respondents will be those who choose to click through AND complete the survey. That's not gonna be anywhere near random, even within the sampling frame.
Again, is that worse than most opinion polls?

User avatar
Dason
Posts: 1309
Joined: Wed Dec 02, 2009 7:06 am UTC
Location: ~/

Re: Statistics Research Paper

Postby Dason » Mon May 03, 2010 1:35 pm UTC

the tree wrote:
Velifer wrote:The population in the survey frame isn't particularly well-defined (anyone who sees his link) and the respondents will be those who choose to click through AND complete the survey. That's not gonna be anywhere near random, even within the sampling frame.
Again, is that worse than most opinion polls?

No, but most opinion polls shouldn't be used for serious research.
double epsilon = -.0000001;

gorcee
Posts: 1501
Joined: Sun Jul 13, 2008 3:14 am UTC

Re: Statistics Research Paper

Postby gorcee » Tue May 04, 2010 2:57 pm UTC

Dason wrote:
the tree wrote:
Velifer wrote:The population in the survey frame isn't particularly well-defined (anyone who sees his link) and the respondents will be those who choose to click through AND complete the survey. That's not gonna be anywhere near random, even within the sampling frame.
Again, is that worse than most opinion polls?

No, but most opinion polls shouldn't be used for serious research.


And yet...

User avatar
Kizyr
Posts: 2070
Joined: Wed Nov 15, 2006 4:16 am UTC
Location: Virginia
Contact:

Re: Statistics Research Paper

Postby Kizyr » Tue May 04, 2010 3:04 pm UTC

This is also a minor statistics project. The results aren't exactly going to be published in some major policy journal. If this were something conducted by, say, Pew or Gallup, we'd expect the method to be a bit more robust. If this is just one individual who doesn't have a ton of resources, writing solely for the purpose of getting an assignment done, it's no big deal.

The OP should add a section on the limitations of his/her study, though, introduced by any sample bias. Other than that, go ahead and use the results to apply what you've learned in class--that's the purpose of the assignment, not to find some grand publishable conclusion. KF
~Kizyr
Image

User avatar
Cleverbeans
Posts: 1378
Joined: Wed Mar 26, 2008 1:16 pm UTC

Re: Statistics Research Paper

Postby Cleverbeans » Tue May 04, 2010 3:15 pm UTC

Also, once you read the survey it's hard to complain to much about the sample bias when the questions are leading. It reminds me of the survey I got when my senator's office called and the machine didn't know how to understand my response when I said answered the question "Do you think the government taxes you to much to support bloated social programs?" and when I said "no" the machine couldn't figure out what I meant after asking three times it just gave up and skipped the question. Good times.
"Labor is prior to, and independent of, capital. Capital is only the fruit of labor, and could never have existed if labor had not first existed. Labor is the superior of capital, and deserves much the higher consideration." - Abraham Lincoln

User avatar
Velifer
Posts: 1132
Joined: Wed Dec 26, 2007 4:05 pm UTC
Location: 40ºN, 83ºW

Re: Statistics Research Paper

Postby Velifer » Wed May 05, 2010 11:33 am UTC

the tree wrote:Again, is that worse than most opinion polls?


Yes. Opinion polling starts with a defined sample frame, and has methods in place for non-response. George Gallup's private army of data thugs will impute the hell out of you if you don't answer.

Kizyr has it right. Play with this and learn a bit, but don't think that it's a good way to collect data.
Time flies like an arrow, fruit flies have nothing to lose but their chains -Marx

Sourire
Posts: 334
Joined: Mon May 25, 2009 3:11 pm UTC

Re: Statistics Research Paper

Postby Sourire » Thu May 06, 2010 6:27 pm UTC

Maybe I'm giving the OP too much credit, but is it possible that the very point of placing the survey here is to demonstrate the bias that so many have expressed concern about?

It just seems to me like we don't really have all the facts of what "minor statistics project" is.
Emi: Let the urge take you on a magic coaster ride of innuendo!

Kewangji: The universe is having an orgasm. Right now.

User avatar
Velifer
Posts: 1132
Joined: Wed Dec 26, 2007 4:05 pm UTC
Location: 40ºN, 83ºW

Re: Statistics Research Paper

Postby Velifer » Fri May 07, 2010 1:44 pm UTC

Sourire wrote:It just seems to me like we don't really have all the facts of what "minor statistics project" is.

I just looked it up:
"The Frequency and Distribution of Forum Posts resulting from Asking Bad Questions, Doctoral Dissertation by A. Magritte, Candidate for Munroe Endowed Chair in Nerd Sniping, Christopher Newport University."

...damn.
Time flies like an arrow, fruit flies have nothing to lose but their chains -Marx

User avatar
the tree
Posts: 801
Joined: Mon Apr 02, 2007 6:23 pm UTC
Location: Behind you

Re: Statistics Research Paper

Postby the tree » Fri May 07, 2010 4:08 pm UTC

Sourire wrote:Maybe I'm giving the OP too much credit, but is it possible that the very point of placing the survey here is to demonstrate the bias that so many have expressed concern about?
That could be the case, they could be posting the same survey in a variety of different forums. Does the googlewhatsit allow you to collect referral data? There could be some interesting results to be gained from that.

User avatar
Kizyr
Posts: 2070
Joined: Wed Nov 15, 2006 4:16 am UTC
Location: Virginia
Contact:

Re: Statistics Research Paper

Postby Kizyr » Fri May 07, 2010 6:12 pm UTC

Velifer wrote:I just looked it up:
"The Frequency and Distribution of Forum Posts resulting from Asking Bad Questions, Doctoral Dissertation by A. Magritte, Candidate for Munroe Endowed Chair in Nerd Sniping, Christopher Newport University."

Gentlemen, I think we've been had.
Damn you, psych majors! KF
~Kizyr
Image

User avatar
OverBored
Posts: 284
Joined: Mon Dec 10, 2007 7:39 pm UTC

Re: Statistics Research Paper

Postby OverBored » Fri May 07, 2010 8:08 pm UTC

Couldn't help but think of some SMBC
G4!!

Grob FTW,

Hello. Smithers. You're. Quite good. At. Turning. Me. On.

Magritte's Apple
Posts: 9
Joined: Wed Jun 18, 2008 3:38 am UTC

Re: Statistics Research Paper

Postby Magritte's Apple » Sat May 15, 2010 2:07 pm UTC

Hey guys!

Okay, I'm already preparing to write a very long explanation of sample bias.

xkcd is certainly not the only place I've placed this survey

and I am indeed a psych major. :wink:

The teacher just wants to see "data" and "conclusion" and then it's fine so long as we explain and show our understanding of "Yes, well, sample bias and volunteer bias running rampant and some of the questions may have been leading" etcetera. As I'm aware some are...It's just a minor exercise. No publishing or anything of that sort.

Though someone did bring up a good point about them not all being mandatory to answer. I should perhaps fix this. Though the due date for the collection of said data is sometime next week and thus far everyone to take it has answered all the questions with only something like 3 exceptions, which I just won't use.

Thanks for pointing things out, though! I will keep them in mind for next time.

You also make a good point about moving it to the General Forum, though you can see by my lack of posts that I'm not actually on the forums often enough to know haha. At that, I've gotten a good amount of samples already.

tesseraktik
Posts: 218
Joined: Fri Nov 06, 2009 9:14 pm UTC
Location: Stockholm, Sweden

Re: Statistics Research Paper

Postby tesseraktik » Wed Jun 09, 2010 6:49 am UTC

Might I ask: Is this poll about our political views as they pertain to our own respective countries, or as they pertain to the U.S.? One can certainly be for/against one's own country's corporate bail-outs, but against/for the U.S. bail-outs, and one can consider one's own country's taxes to be too high/low while considering taxes in the U.S. to be too low/high.
ni'o mi nelci le zirpu sovmabrnornitorinku
Spoiler:
++$_ wrote:What's a "degree"?

EDIT: I looked it up on Wikipedia. Apparently it's some ancient Babylonian unit for angles :/

User avatar
4=5
Posts: 2073
Joined: Sat Apr 28, 2007 3:02 am UTC

Re: Statistics Research Paper

Postby 4=5 » Wed Jun 09, 2010 2:10 pm UTC

please DO NOT fix the ability to skip questions, some questions the answer is "it depends" and the only way I can show that is by skipping


Return to “Mathematics”

Who is online

Users browsing this forum: No registered users and 5 guests