[Openspace] re: Spatial Empirical Bayes with a lot of zeros

[Wayne E Thogmartin]

It stands to reason that you would have (somewhat) fewer areas of zero
prevalence, as the spatial smoothing would predict some level of nonzero
prevalence for areas adjacent to areas of nonzero prevalence.  The
predicted prevalence would be pulled up away from the observed zero
prevalence to some level of predicted prevalence less than the neighboring
nonzero prevalences.

Wayne E. Thogmartin, PhD
USGS Upper Midwest Environmental Sciences Center
2630 Fanta Reed Road
La Crosse, WI 54603
608-781-6309 (off)
608-783-6066 (fax)
wthogmartin at usgs.gov
http://www.umesc.er.usgs.gov/terrestrial/migratory_birds/bird_conservation.html


The contents of this message are mine personally and do not reflect any
portion of the US government.

"A recurrent theme in the history of science is that novel claims conjoined
with a paucity of data inevitably attract the attention of statisticians,
much in the manner that offal attracts flies."  W.M. Schaffer


|---------+------------------------------------->
|         |           openspace-request at sal.agec|
|         |           on.uiuc.edu               |
|         |           Sent by:                  |
|         |           openspace-bounces at sal.agec|
|         |           on.uiuc.edu               |
|         |                                     |
|         |                                     |
|         |           11/24/2004 12:00 PM       |
|         |           Please respond to         |
|         |           openspace                 |
|---------+------------------------------------->
  >------------------------------------------------------------------------------------------------------------------------|
  |                                                                                                                        |
  |       To:       openspace at sal.agecon.uiuc.edu                                                                          |
  |       cc:                                                                                                              |
  |       Subject:  Openspace Digest, Vol 21, Issue 6                                                                      |
  >------------------------------------------------------------------------------------------------------------------------|




Send Openspace mailing list submissions to
             openspace at sal.agecon.uiuc.edu

To subscribe or unsubscribe via the World Wide Web, visit
             http://sal.agecon.uiuc.edu/mailman/listinfo/openspace
or, via email, send a message with subject or body 'help' to
             openspace-request at sal.agecon.uiuc.edu

You can reach the person managing the list at
             openspace-owner at sal.agecon.uiuc.edu

When replying, please edit your Subject line so it is more specific
than "Re: Contents of Openspace digest..."


Today's Topics:

   1. Spatial Empirical Bayes with a lot of zeros


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

Date: Tue, 23 Nov 2004 15:49:01 -0500
From: cbautista at hivresearch.org
To: openspace at agec221.agecon.uiuc.edu
Subject: [Openspace] Spatial Empirical Bayes with a lot of zeros
Message-ID:
<OFA0F717FD.3DC51BB5-ON85256F55.0070D01A-85256F55.00725A37 at hivresearch.org>
In-Reply-To: <e3595fc4.3950c6e5.81e4e00 at expms4.cites.uiuc.edu>
Content-Type: text/plain; charset=us-ascii
MIME-Version: 1.0
Precedence: list
Message: 1


I have a study where the number of areas is 300, and for each area I have
calculated the prevalence or proportion of a disease.

Almost 70% of the total of areas in this study the prevalence (rate) is
equal to 0.


Using GeoDa (excellent software!!) with the option of Spatial Empirical
Bayes I would like to create a Bayesian smoothed prevalences map. After
using this option and having created the respective weight file, I have
saved "the rates" for each area. However, I have observed that now there
are less areas with zero prevalences.

a) Am I not sure about the spatial empirical bayes method is a good option
for my study?

b) Is there any recomendation about what to do when there are many zeros?

c) Is there another way of creating a Bayesian smoothed prevalence map?

Any comments please feel free....

Thanks in advance,

/Christian






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

_______________________________________________
Openspace mailing list
Openspace at sal.agecon.uiuc.edu
http://sal.agecon.uiuc.edu/mailman/listinfo/openspace


End of Openspace Digest, Vol 21, Issue 6
****************************************