Mapping the Landscape of Aging

How were Clients Mapped?

In 2004 Jefferson Area Board of Aging served more than 1700 different clients. Many of the clients received multiple services. These services are described in detail at the JABA websites Services Page.

Client services are tracked by the adminstrators of JABA and reported to the Virginia Office of Aging and other Aging administrative offices. The classes of services so reported include: Case Management Assistance, Access and Acquisition, Disease Prevention and Health Promotion, Health Assessment, Health Services and Medication Management, Adult Day Care, the Cool Aid Program, Home Delivered Meals, Home Safety, Home Care, Senior Center Clients, Transportation, and Congregate Volunteers. These are described in the Client Mapping page.

Jefferson Area Board of Aging worked collaborated with the Mapping Project to make the spatial distribution of the clients available while insuring the confidentiality of the clients. A list of client addresses with an unique "index number" unrelated to any administrative code was provided the researchers. No names, administrative numbers or other codes by which a client might be identified were released by JABA to the research team. The addresses were parsed to be consistent in format with the form desired by an address matching, geocoding vendor. These addresses along with other addresses of facilities and services were sent to Geographic Data Technology Services, (recently acquired by TeleAtlas).

The geocoding service returned a XY location in Longitude/Latitude coordinates for each address it could successfully match against its address base. When a direct match was not possible, the XY location of a Zipcode zone or a political jursidiction was provided. The type of "match" and the type of coordinate was reported. The "index number" remained part of the record. These XY locations were then stripped of the addresses. The "index number" was used as the key to match the XY locations with the administratively collected "type of service" information for each client record. Again, no information about the individual clients were released to the researchers.

To further protect the confidentiaity of the clients, the geocoded location of the client was purturbed by a random distance from the actual location so that no point would be within one-half block of its actual location. Consequently the pattern of clients can be analyzed, but the specific location of any client cannot be derived.

About 98% of the clients could be placed on the map. The remaining clients had P.O. Boxes or incorrect street addresses. The methodological note discusses the efficiency of the address coding and matching.