iid distribution problem

classic Classic list List threaded Threaded
25 messages Options
12
Reply | Threaded
Open this post in threaded view
|

Re: mixed models

drfg2008
Both approaches have their advantages and disadvantages. The Milgram experiment is based not on a realistic situtation, because the university as an environment and science, with its high reputation are likely to have had a significant influence on the results. Moreover, the interpretation of the result is of great controversy. Of course, opinions differ from actions.  The present vignette analysis is about attitudes and opinions, and these are not exactly 1:1 translated into actions. What actions will ultimately follow, depends on many other factors, which in large part simply can not be simulated experimentally.

But what has that to do with the question if SPSS mixed models is the right approach to solve this specific statistical problem.

Frank
Dr. Frank Gaeth

Reply | Threaded
Open this post in threaded view
|

Re: mixed models

Bruce Weaver
Administrator
It has nothing to do with the discussion about choosing an appropriate statistical model.  That is why I said the following at the top of my post:

Off-Topic

This comment is not about the statistics.  Rather, it is concerned with what conclusions one can draw from studies that have people read vignettes and then indicate how they would behave or respond in that situation.



drfg2008 wrote
Both approaches have their advantages and disadvantages. The Milgram experiment is based not on a realistic situtation, because the university as an environment and science, with its high reputation are likely to have had a significant influence on the results. Moreover, the interpretation of the result is of great controversy. Of course, opinions differ from actions.  The present vignette analysis is about attitudes and opinions, and these are not exactly 1:1 translated into actions. What actions will ultimately follow, depends on many other factors, which in large part simply can not be simulated experimentally.

But what has that to do with the question if SPSS mixed models is the right approach to solve this specific statistical problem.

Frank
--
Bruce Weaver
bweaver@lakeheadu.ca
http://sites.google.com/a/lakeheadu.ca/bweaver/

"When all else fails, RTFM."

PLEASE NOTE THE FOLLOWING: 
1. My Hotmail account is not monitored regularly. To send me an e-mail, please use the address shown above.
2. The SPSSX Discussion forum on Nabble is no longer linked to the SPSSX-L listserv administered by UGA (https://listserv.uga.edu/).
Reply | Threaded
Open this post in threaded view
|

Re: mixed models

Art Kendall
In reply to this post by drfg2008
I concur that it is well established that actions are weakly related to intentions, attitudes, and values.

The objection to the use of vignettes used an example of people's reactions to a single vignette.

In many uses of vignettes, elements of the vignette are changed to operationalize levels of independent (design) variables.  They can be conceptualized as stimuli that are deliberately varied.  Sometimes there is a simple dichotomy with race or gender being the element of the vignette that is changed.  Preferences/choices can be analyzed, for example, by changing price, quality, and speed of delivery. etc. (Remember the old quip. "Cheap, fast, good. Pick two.")

Suggesting a statistical approach depends very heavily on what questions the research is designed to answer. What questions are you approaching?

Please describe your design in more detail. Different dialects of statistics vary in their use of terminology so please try to clarify your question. I for one do not have a clear idea of what the statistical problem is.

How do the 16 vignettes vary?  Are they designed to vary along one dimension? Two dimensions? Three? Four?

Are they used as levels of repeated measures (within subject/respondent)  or are different people presented with different vignettes?


How many groups of respondents do you have?  What are the sizes of the groups?
What is the design by which the vignettes were constructed?

Art Kendall
Social Research Consultants

On 9/18/2011 12:14 PM, drfg2008 wrote:
Both approaches have their advantages and disadvantages. The Milgram
experiment is based not on a realistic situtation, because the university as
an environment and science, with its high reputation are likely to have had
a significant influence on the results. Moreover, the interpretation of the
result is of great controversy. Of course, opinions differ from actions.
The present vignette analysis is about attitudes and opinions, and these are
not exactly 1:1 translated into actions. What actions will ultimately
follow, depends on many other factors, which in large part simply can not be
simulated experimentally.

But what has that to do with the question if SPSS mixed models is the right
approach to solve this specific statistical problem.

Frank

-----
Dr. Frank Gaeth
FU-Berlin

--
View this message in context: http://spssx-discussion.1045642.n5.nabble.com/iid-distribution-problem-tp4746052p4816329.html
Sent from the SPSSX Discussion mailing list archive at Nabble.com.

=====================
To manage your subscription to SPSSX-L, send a message to
[hidden email] (not to SPSSX-L), with no body text except the
command. To leave the list, send the command
SIGNOFF SPSSX-L
For a list of commands to manage subscriptions, send the command
INFO REFCARD

===================== To manage your subscription to SPSSX-L, send a message to [hidden email] (not to SPSSX-L), with no body text except the command. To leave the list, send the command SIGNOFF SPSSX-L For a list of commands to manage subscriptions, send the command INFO REFCARD
Art Kendall
Social Research Consultants
Reply | Threaded
Open this post in threaded view
|

Re: mixed models

Mike
In reply to this post by drfg2008
On Sunday, September 18, 2011 12:14 PM,  "drfg2008" wrote:
> Both approaches have their advantages and disadvantages. The Milgram
> experiment is based not on a realistic situtation, because the university as
> an environment and science, with its high reputation are likely to have had
> a significant influence on the results.

Of course, this is nonsense.  Jerry Burger of the "high reputation"
institution of Santa Clara University successfully conducted a partial
replication of Miglram's conditions and found very similar results.
Burger had the research published in the American Psychologist and
a copy can be obtained here:
https://www.scu.edu/cas/psychology/faculty/upload/Replicating-Milgrampdf.pdf

Getting IRB approval for what some consider the most unethical
psychology experiment ever done was difficult (Milgram and Zimbardo
grew up in Queens, NY and were friends;  Milgram told Zimbardo
after the Stanford Prison Experiment that he was glad Zimbardo
could take some of the heat off of him for having done one of the most
unethical psychological experiments).  Burger talks a little about the process
for getting IRB approval in an article in the Association for Psychological
Science's (APS) Observer; see:
http://www.psychologicalscience.org/observer/getArticle.cfm?id=2264

As for the external validity of Milgram's results, I would suggest taking
a look at Alex Gibney's documentary "The Human Behavior Experiments"
which focus on the classic Milgram, Bystander Apathy, and Stanford
Prison experiments as well as real world counterparts; see:
http://www.imdb.com/title/tt0822813/
This documentary is presented a couple of times a year on the Sundance
channel (portions are also available on the YouTube).

-Mike Palij
New York University
[hidden email]

=====================
To manage your subscription to SPSSX-L, send a message to
[hidden email] (not to SPSSX-L), with no body text except the
command. To leave the list, send the command
SIGNOFF SPSSX-L
For a list of commands to manage subscriptions, send the command
INFO REFCARD
Reply | Threaded
Open this post in threaded view
|

Re: mixed models

drfg2008
Of course, this is nonsense.  Jerry Burger of the "high reputation"
institution of Santa Clara University successfully conducted a partial
replication of Miglram's conditions and found very similar results.


Well, if you replicate an experiment in the same environment under similar conditions, you will most likely come up with the same results. However, I'm glad to hear that the "Germans-are-different" thesis has been proven wrong. ;-)
Dr. Frank Gaeth

12