bimodal data

Hi Simone,
It depends on what you want from the data ...
A bimodal distribution can typically be modelled with a mixture
distribution, where the assumption is that each of the modes corresponds to
an underlying process generating those data.
hth, ingmar

  Hi,
 
Does anybody have a good tip of how to treat bimodal data to perform
statistical analyses? My data set ranges from -1 to 1 (any values are
posssible in between) and most data are either close to -1 or close to 1. They
are the results of a two choice experiment where individuals could choose more
than once in either direction and scores were calculated.
 
Simone



Simone Immler
University of Sheffield
Dep. Animal & Plant Sciences
Alfred Denny Building
Western Bank
Sheffield S10 2TN, UK


---------------------------------

[[alternative HTML version deleted]]

______________________________________________
R-help at stat.math.ethz.ch mailing list
https://stat.ethz.ch/mailman/listinfo/r-help
PLEASE do read the posting guide! http://www.R-project.org/posting-guide.html

    Hi,
  Does anybody have a good tip of how to treat bimodal data to
perform statistical analyses? My data set ranges from -1 to 1
(any values are posssible in between) and most data are either
close to -1 or close to 1. They are the results of a two choice
experiment where individuals could choose more than once in
either direction and scores were calculated.
  Simone

Frankly, I suspect you will do better to analyse at the level of
the original choices, rather than the final "scores", for two
reasons:

1. It will probably be easier, more straightforward, and more
tractable to set up a statistical model for the binary choices;

Distributions which (as I infer from your description) are highly
"U-shaped" are tricky to model, and do not lend themselves to
the conventional types of analysis at all well.

2. Because "scores were calculated" you have probably lost
useful information.

One issue that may be informative (which you do not specify) is
whether the number of choices made varies from subject to subject
and, if so, how many choices did each subject make.

Hoping this helps, and please come back with more detail if it
might help to make progress.

Best wishes,
Ted.


--------------------------------------------------------------------
E-Mail: (Ted Harding) <Ted.Harding at nessie.mcc.ac.uk>
Fax-to-email: +44 (0)870 094 0861
Date: 02-Dec-05                                       Time: 13:35:55
------------------------------ XFMail ------------------------------

Your problem sounds like it could be modeled with logistic regression
whereby the propensity for one result or another is "linked" to the factors
that control it.  Logistic regressions are a special case of generalized
linear models.  Look at ?glm



Charles Annis, P.E.

Charles.Annis at StatisticalEngineering.com
phone: 561-352-9699
eFax:  614-455-3265
http://www.StatisticalEngineering.com
 

-----Original Message-----
From: r-help-bounces at stat.math.ethz.ch
[mailto:r-help-bounces at stat.math.ethz.ch] On Behalf Of Simone Immler
Sent: Friday, December 02, 2005 7:43 AM
To: r-help at stat.math.ethz.ch
Subject: [R] bimodal data

    Hi,
   
  Does anybody have a good tip of how to treat bimodal data to perform
statistical analyses? My data set ranges from -1 to 1 (any values are
posssible in between) and most data are either close to -1 or close to 1.
They are the results of a two choice experiment where individuals could
choose more than once in either direction and scores were calculated.
   
  Simone



Simone Immler
University of Sheffield
Dep. Animal & Plant Sciences
Alfred Denny Building
Western Bank
Sheffield S10 2TN, UK

		
---------------------------------


______________________________________________
R-help at stat.math.ethz.ch mailing list
https://stat.ethz.ch/mailman/listinfo/r-help
PLEASE do read the posting guide!
http://www.R-project.org/posting-guide.html