From Richard.Hoskins@DOH.WA.GOV Tue Aug  1 11:39:54 2000
Received: from mxu4.u.washington.edu (mxu4.u.washington.edu [140.142.33.8])
	by lists.u.washington.edu (8.9.3+UW00.05/8.9.3+UW99.09) with ESMTP id LAA145798
	for <waphgis@lists.u.washington.edu>; Tue, 1 Aug 2000 11:39:54 -0700
Received: from mail1.doh.wa.gov (mail1.doh.wa.gov [192.230.11.132])
	by mxu4.u.washington.edu (8.9.3+UW00.02/8.9.3+UW99.09) with ESMTP id LAA08860
	for <waphgis@u.washington.edu>; Tue, 1 Aug 2000 11:39:53 -0700
Received: by mail1.doh.wa.gov with Internet Mail Service (5.5.2652.35)
	id <PHNT5LA6>; Tue, 1 Aug 2000 11:35:48 -0700
Message-ID: <213895AE8149D3118D9500508B0F20A0013854BB@mail2.doh.wa.gov>
From: "Hoskins, Richard" <Richard.Hoskins@DOH.WA.GOV>
To: "'waphgis@u.washington.edu'" <waphgis@u.washington.edu>
Subject: WAPHGIS:  Jenk's Method
Date: Tue, 1 Aug 2000 11:35:36 -0700 
MIME-Version: 1.0
X-Mailer: Internet Mail Service (5.5.2652.35)
Content-Type: text/plain;
	charset="iso-8859-1"

I just finished reading the Smith paper referred to below. Pretty
intresting. I'd be interested in finding some more recent work on this topic
of selecting class intervals. Is anyone familiar with methods in which
classes can be displayed that are "conditioned" on the values of other
covariates? Examples? 

Dick Hoskins

-----Original Message-----
From: fpb01@health.state.ny.us [mailto:fpb01@health.state.ny.us]
Sent: Monday, July 24, 2000 8:52 AM
To: waphgis@u.washington.edu
Subject: Jenk's Method


For decades cartographers have drawn histograms of data distributions, seen
how sometimes the data points form clusters, and selected these "natural
breaks" as their class breaks. Jenk's Method is the formal mathematical
expression of this technique. It defines categories that have the highest
within-category similarity and between-category dissimilarity.

Jenk's Method works well for data with multimodal and highly skewed
distributions. It does not put dissimilar values together in the same
category, which occasionally happens when using quantiles.

See Richard Smith, "Comparing Traditional Methods for Selecting Class
Intervals on Choropleth Maps", The Professional Geographer 38(1986): 62-67.

If you make a map in either MapInfo or ArcView and choose "natural breaks",
you are using Jenk's Method.

Frank Boscoe
NYS Department of Health

.