From lwaller@sph.emory.edu Mon Apr 17 06:42:24 2000
Received: from mxu4.u.washington.edu (mxu4.u.washington.edu [140.142.33.8])
	by lists.u.washington.edu (8.9.3+UW99.09/8.9.3+UW99.09) with ESMTP id GAA33286
	for <waphgis@lists.u.washington.edu>; Mon, 17 Apr 2000 06:42:22 -0700
Received: from gator.sph.emory.edu (gator.sph.emory.edu [170.140.4.2])
	by mxu4.u.washington.edu (8.9.3+UW00.02/8.9.3+UW99.09) with ESMTP id GAA01548
	for <waphgis@u.washington.edu>; Mon, 17 Apr 2000 06:42:22 -0700
Received: from viper.sph.emory.edu (root@viper.sph.emory.edu [170.140.4.1])
	by gator.sph.emory.edu (8.10.1/8.10.1) with ESMTP id e3HDgHr00358
	for <waphgis@u.washington.edu>; Mon, 17 Apr 2000 09:42:18 -0400 (EDT)
Received: from sph.emory.edu (squid.sph.emory.edu [170.140.4.9])
	by viper.sph.emory.edu (8.10.1/8.10.1) with ESMTP id e3HDgG906098
	for <waphgis@u.washington.edu>; Mon, 17 Apr 2000 09:42:16 -0400 (EDT)
Sender: lwaller@sph.emory.edu
Message-ID: <38FB14B4.BDE489AB@sph.emory.edu>
Date: Mon, 17 Apr 2000 09:42:12 -0400
From: Lance Waller <lwaller@sph.emory.edu>
X-Mailer: Mozilla 4.7 [en] (X11; U; SunOS 5.7 sun4u)
X-Accept-Language: en
MIME-Version: 1.0
To: waphgis@u.washington.edu
Subject: Re: WAPHGIS:  rural versus urban
References: <200004151521.LAA19848@smtpsrv1.isis.unc.edu>
Content-Type: text/plain; charset=us-ascii
Content-Transfer-Encoding: 7bit

Thanks for the information on the Beale codes, very
interesting.

I have a brief comment and question for the list regarding
possibly extending to (US) census tracts.

I've analyzed some census tract level data from central
NY and have some impressions about tract data that I would
like the list to comment on as to their generality.

Some background (as I understand it, corrections encouraged):

- Tracts were originally designed to have roughly the same
  population size in each (about 3000 people or so).  Naturally,
  this results in geographically small tracts in areas with
  higher population density, and geographically large tracts
  in areas with lower pop'n density.

Now, (my data is 1980 census data) I naively expected to
find higher pop'n sizes in urban areas over time, but what
I find is much higher *variability* (big AND small numbers) 
in pop'n size in "urban"
tracts than in "rural" tracts (I suppose due to rezoning, more
mobile population etc).  The result is that the lowest pop'n
sizes often occur in "urban" tracts (in my case, a pop'n size
of 3 people in one tract in Syracuse), while the "rural" 
tracts stay about the same size (less mobile and larger 
geographically, so people have to move farther to get out
of the tract).  Some rural tracts have small pop'n sizes,
but most of the small ones seem to be in the cities in my
data set.

This is of interest to me since some spatial smoothing procedures
for rates argue about small numerators (pop'n sizes) in rural
tracts when the problem is actually often worse in urban tracts.

Anyone else have this experience?

(I expect ZIPs are a bit more "stable" (ironic use of the word as
ZIPs can change at any time) in pop'n size since they
are reassigned if they get too small or large (pop'n wise)).

Lance Waller
Dept of Biostatistics
Rollins School of Public Health
Emory University 
Atlanta, GA
.