From g.lauer@geog.canterbury.ac.nz Mon Apr 19 15:34:25 1999 Received: from mxu3.u.washington.edu (mxu3.u.washington.edu [140.142.33.7]) by lists.u.washington.edu (8.9.3+UW99.02/8.9.3+UW99.01) with ESMTP id PAA30544 for ; Mon, 19 Apr 1999 15:34:21 -0700 Received: from smtp2.ihug.co.nz (tk2.ihug.co.nz [203.29.160.14]) by mxu3.u.washington.edu (8.9.3+UW99.02/8.9.3+UW99.01) with ESMTP id PAA11249 for ; Mon, 19 Apr 1999 15:34:19 -0700 Received: from [207.214.13.173] (p46-max4.chc.ihug.co.nz [207.214.13.173]) by smtp2.ihug.co.nz (8.8.8/8.8.8) with ESMTP id KAA28316 for ; Tue, 20 Apr 1999 10:34:14 +1200 Mime-Version: 1.0 Content-Type: text/plain; charset="us-ascii" X-Sender: gml30@pophost.canterbury.ac.nz (Unverified) Message-Id: In-Reply-To: <002801be8999$111f0600$7b790518@olmpi1.wa.home.com> References: Date: Tue, 20 Apr 1999 01:04:48 +1200 To: waphgis@u.washington.edu From: Greg Lauer Subject: Re: WAPHGIS: NZ hospital admin project Thanks for your comments For the project I am undertaking I need to have the data at meshblock level (~50 housholds). The agency that collects the data does not supply the data to this level. They do at higher level (CensusArea Unit). Problem is the that the accuracy rate is unknown, and could be less than 60%. This is mainly due to the lack of a geocoding standards in NZ. Thus I need individual address data. Once aggragated to meshblock level the addresses will be stripped. Then the data will be further utilised. I totaly understand the need to preserve confidentiality/privacy. Wouldnt want someone getting hold of my health data. Sort of between a rock and a hard place though:-) I am awaiting a reply from the regional ethics committee. Fingers crossed...... Cheers Greg >Greg: In my state if we are doing a research project it is pretty routine >to get an IRB (institutional review board) approval for using address data. >If you have boundaries for the census unit and you can geocode down to the >census unit then there is no reason to go down to street level. If you can't >figure out what census unit they are in but you can get to the street level >then you will need to do a point/polygon query in your GIS to get that info. >I assume you have population numbers for the census units and you want to >calculate rates? or what? Our IRB requires that we do not share the data or >identifiers, or remove names. As you know calculating rates with small units >like this can be a problem, so we adjust rates using empirical Bayesian >smoothing methods or Tukey headbanging, the method used in the NCHS National >Mortality Atlas. Greg Lauer Department of Geography University of Canterbury PO Box 36348 Merivale Christchurch, New Zealand .