|
Korean Job Discussion Forums
"The Internet's Meeting Place for ESL/EFL Teachers from Around the World!"
|
| View previous topic :: View next topic |
| Author |
Message |
Woland
Joined: 10 May 2006
Location: Seoul
|
 Posted: Fri Dec 01, 2006 4:45 am Post subject: Psychotics favored Bush |
 |
|
This study both amuses me and makes me sad at the same time:
http://www.ctnow.com/custom/nmm/newhavenadvocate/hce-nha-1123-nh48bushbash48.artnov23,0,1695911.story
"The thesis draws on a survey of 69 psychiatric outpatients in three Connecticut locations during the 2004 presidential election. Lohse�s study, backed by SCSU Psychology professor Jaak Rakfeldt and statistician Misty Ginacola, found a correlation between the severity of a person�s psychosis and their preferences for president: The more psychotic the voter, the more likely they were to vote for Bush."
Yes, it's a limited N, but the results are what they are. Please note that the author is careful to say that not all Bush supporters are psychotic and does offer a reasonable explanation for why the results turned out this way. Note, too, that the author of the study describes himself as a Reaganite. |
|
| Back to top |
|
 |
Ya-ta Boy
Joined: 16 Jan 2003
Location: Established in 1994
|
| Posted: Fri Dec 01, 2006 6:40 am Post subject: |
|
|
| Quote: |
| Please note that the author is careful to say that not all Bush supporters are |
Can the author present any credible evidence that Bush supporters are not psychotics? Or is it just wishful thinking or a wild guess? |
|
| Back to top |
|
|
Satori
Joined: 09 Dec 2005
Location: Above it all
|
| Posted: Fri Dec 01, 2006 10:18 am Post subject: |
|
|
| Im anti Bush, but Im also anti bogus science. 69 subjects is not a survey, it means nothing. |
|
| Back to top |
|
|
ddeubel
Joined: 20 Jul 2005
|
| Posted: Fri Dec 01, 2006 1:00 pm Post subject: |
|
|
Don't you just hate all these scientists, lab coat types, ratings rats who spend so much time and energy just to confirm the obvious???
Seems too much of all research and survey is such. Bad scientific methodology. Putting the thesis before the facts and making Heisenberg look good. Like doing those puzzles when already knowing what the picture looks like on the box.
I think a little more inductive reasoning is called for.....also Occam's Razor would seem to apply here as a general principal.
DD |
|
| Back to top |
|
|
mnhnhyouh
Joined: 21 Nov 2006
Location: The Middle Kingdom
|
| Posted: Fri Dec 01, 2006 1:14 pm Post subject: |
|
|
Outside of mathematics, inductive reasoning is not valid.
h |
|
| Back to top |
|
|
Roch
Joined: 24 Apr 2003
Location: Seoul
|
| Posted: Mon Dec 04, 2006 7:12 pm Post subject: |
|
|
| Satori wrote: |
| Im anti Bush, but Im also anti bogus science. 69 subjects is not a survey, it means nothing. |
Word. |
|
| Back to top |
|
|
Woland
Joined: 10 May 2006
Location: Seoul
|
| Posted: Tue Dec 05, 2006 12:01 am Post subject: |
|
|
| Roch wrote: |
| Satori wrote: |
| Im anti Bush, but Im also anti bogus science. 69 subjects is not a survey, it means nothing. |
Word. |
I posted the original because I saw the humor in it (which is probably also why it got picked up in the press). I added the dislcaimers because I thought there would be some people here would misread this as a claim that all Bush supporters are psychotic and get all over me. But the claims that this is bad science disappoint me. There's plenty of evidence that this is good science, but I'll confess a lot of it is indirect as a result of this being a piece of journalistic writing.
As Charles Bazerman has pointed out, the genre differences in academic scientific and journalistic science writiing often lead to distortions in what gets presented in the press to the public about scientific studies. Journalistic writing is interested in eye-catching results, rather than the more limited claims common in academic writing. Journalistic writing is also not generally interested in the details of methodology that scientists are because the public isn't interested in them. Their loss.
Reading the full article, it appears that this research was done for the author's master's thesis at Southern Connecticutt State University. In order for it to pass, he had to go through approval of his proposal (whether by advisor alone or by committee), human subjects approval by a university committee, and a final defense with two readers other than his advisor. One of those readers is generally required to be external to the whole process, though I don't know if this is the case here. To get past all those professional readers and be bad science is unlikely.
Beyond this, the fact that word of the research has reached the public realm via the press suggests that it has been published somewhere, which means it was vetted again by two outside, external readers, as well as the editor of the journal it appeared in. There's no mention of a particular journal in the article, so this may not be the case. The newspaper is a local one in New Haven, so it may be that the reporter heard of it locally. Still, the fact that the both the supervising professor and the statistician involved are both willing to be publicly identified with the work again suggests that the science is good.
Satori's complaint that the N of 69 is a sign that it is bad science is simplistic and suggests a lack of understanding of how survey research works. It also points to an area where the article doesn't really give us enough information about method. I have tried with my comments above to suggest things about journalistic writing that result in this lack, but let me address this issue directly here.
In survey research, one ideally would survey every member of the population that one is trying to find out about. This is rarely possible, and researchers use a sample population instead. If the sample is a proper representative of the total population that one wishes to examine, a very small sample can accurately represent a very large population. For example, polling agencies in the US regularly make quite accurate predictions about the voting behaviors of populations in the millions based on properly determined samples of as few as 500 people. Again, I would assume that one of the purposes of consulting a statistician in the research process would be to look at things like sampling in the design to ensure that it is done properly. I would also assume that the statistician approved of the sample because she is allowing her name to be used publically in reporting the results. Her reputation among her peers is put on the line by things like that.
What's missing from the article that we could use is information about what the target population is - is it the population of mentally ill people involved in the voting support study, the population of mentally ill people in Connecticutt, or the whole US. Again, I assume that this is specified in the original report, as well as how the sample represents it (and how that was determined). With whatever target population, I again assume that the statistician approved the sample size. Sixty-nine psychotics could be a perfectly good sample, and again, other evidence here suggests it is. If there is any sampling bias (sometimes inescapable in the real world), this would also be reported in the limitations section in the original article/thesis, but it is the kind of thing that doesn't make it into the press because it isn't interesting to readers, although it is essential to understanding what the results mean.
I'm also assuming that the standard error of measure of the instrument is reported. This usually makes it into press reports of surveys as "with a margin of error of plus or minus X percent," but didn't here because they didn't report the level of correlation, another bit of missing info. A weak correlation can be significant if the sample size is right; though with 69 subjects, I'm betting the correlation is moderate to strong.
In short, there's missing information here, a result of the translation of a scientific article in to journalism, but there's substantial indirect evidence that the science is good. Certainly, going simply from the N to a claim that the science is bad is unsupported in general.
Just to give you my bona fides on this, I have my PhD, currently teach the research methods course in my program at my university, have supervised research-based MAs for seven years, and have published my own research. I frequently use surveys in my research, am quite good at their design and validation, including the creation of scales within a larger instrument, and am familiar with issues of sampling.
Last edited by Woland on Thu Dec 07, 2006 6:13 am; edited 1 time in total |
|
| Back to top |
|
|
cbclark4
Joined: 20 Aug 2006
Location: Masan
|
| Posted: Wed Dec 06, 2006 8:06 pm Post subject: |
|
|
I can't see how the ydrew that conclusion, I'm pretty sure that Kerry won the homeless vote hands down.
cbc |
|
| Back to top |
|
|
Big_Bird
Joined: 31 Jan 2003
Location: Sometimes here sometimes there...
|
| Posted: Wed Dec 06, 2006 8:12 pm Post subject: |
|
|
| Ya-ta Boy wrote: |
| Quote: |
| Please note that the author is careful to say that not all Bush supporters are |
Can the author present any credible evidence that Bush supporters are not psychotics? Or is it just wishful thinking or a wild guess? |
 |
|
| Back to top |
|
|
Big_Bird
Joined: 31 Jan 2003
Location: Sometimes here sometimes there...
|
| Posted: Wed Dec 06, 2006 8:15 pm Post subject: |
|
|
Thanks Woland, having studied a whole load of dreary stats as an undergrad, I appreciated your explanation given in your last post.  |
|
| Back to top |
|
|
Woland
Joined: 10 May 2006
Location: Seoul
|
| Posted: Thu Dec 07, 2006 6:45 am Post subject: |
|
|
| cbclark4 wrote: |
I can't see how the ydrew that conclusion, I'm pretty sure that Kerry won the homeless vote hands down.
cbc |
The study didn't look at who the psychotics voted for, although the data is suggestive in that regard. It looked at political views and whether that related to level of psychosis.
The study was correlational. The survey instrument actually contained two different surveys. One was a standard instrument used to measure psychosis. I believe it is even identified by name in the article. The other was a survey of political views. The correlation of results of the two surveys for the subjects was calculated, probably using the Pearson Product Moment Correlation.
In all likelihood, the actual survey contained nothing about Bush. The author is drawing an inference connecting certain political views to Bush. That connection would probably be easy to demonstrate.
**********
Thanks for the kind words, Big Bird. I really think that basic stats should be part of everyone's education. It's unfortunate that math in general, and stats in particular are taught in ways that make people intimidated by them. Part of what I have to do in teaching research methods to teachers is reduce that feeling of intimidation. I want them to see that the math is manageable and understandable and that knowing how you will measure things will help you design your study and also guide you in drawing conclusions from your data.
I've gone back and cleaned up that long post a bit. I am a terrible typist and worse proofreader. |
|
| Back to top |
|
|
Big_Bird
Joined: 31 Jan 2003
Location: Sometimes here sometimes there...
|
| Posted: Fri Dec 08, 2006 6:27 pm Post subject: |
|
|
| Woland wrote: |
Thanks for the kind words, Big Bird. I really think that basic stats should be part of everyone's education. It's unfortunate that math in general, and stats in particular are taught in ways that make people intimidated by them. Part of what I have to do in teaching research methods to teachers is reduce that feeling of intimidation. I want them to see that the math is manageable and understandable and that knowing how you will measure things will help you design your study and also guide you in drawing conclusions from your data.
|
Yes, I agree that having a good understanding of stats and being able to interpret them is a very useful skill that should be more widely taught. I started off doing a maths degree, so I've never been that intimidated, but I did a Research Methods course recently and saw that 90% of the class were sitting there in terror, despite the fact the lecturer was going at snails pace explaining it and giving many useful examples. People just seem to turn off at the mention of maths or stats and believe that they won't understand. He spent 4 weeks covering what would have been covered in less than one hour in one of my undergrad stats courses. Good luck in your endevours to make it more accessible to your students. |
|
| Back to top |
|
|
|
|
You cannot post new topics in this forum
You cannot reply to topics in this forum
You cannot edit your posts in this forum
You cannot delete your posts in this forum
You cannot vote in polls in this forum
|
|
FAQ
Search
Memberlist
Usergroups
Register
Profile
Log in to check your private messages
Log in
