Goss, Jon. 1995. "'We know who you are and we know where you live': The instrumental rationality of geodemographic systems."
As I was reading this article I
could not stop coming up with examples of geodemographics and the problems Goss
discusses; so rather than a summary, I’m going to give some definitions, digestible
examples, and make a few connections based on our reading this semester.
Geodemographics is a form of
statistical analysis of a person and his/her behaviors based on that person’s
identity/demographic and location/geography. Marketers can use this information
to identify new markets or develop better marketing strategies for existing
markets.
This article reminds me of the
methodology section of Christine Pawley’s article "Seeking 'Significance:'
Actual Readers, Specific Reading Communities," where she explains that
with the invention of statistics people began self-identifying using
statistical terms: married, white, female, graduate education, democrat.
This article explains how geodemographics takes those expected and perhaps
obvious statistical categories and creates models by comparing and cross-referencing
many databases of information to create highly customizable marketing
strategies.
Some of the information these
massive data gatherers collect is information we as individuals freely give,
but gets aggregated into a database without our knowledge or consent; for
example, we might sign up for a frequent-shopper card to get discounts, but then
our entire shopping history can be tracked, which provides the retailer
information about us and the potential preferences of our demographic. This
data gathering is a form of surveillance; therefore privacy is at the center of
the geodemographic ethical debate.
Another problem is with the methods
or parameters set by geographers and statisticians to create geographical
information system (GIS) models. Goss explains that, “The related ecological fallacy
is perhaps the most serious technical problem affecting spatial analysis in
GIS. This refers to the erroneous assumption that patterns or relationships
between data observed at an aggregate level of analysis also apply to data at
the level of the individual (or to any lower level of aggregation” (p. 181).
Patterns that seem obvious using GIS might be incorrect because of the “arbitrary
nature of statistical boundaries.”
In this map you can see the
concentration of people with bachelor's degrees by county. The counties that
include Washington, DC; Bloomington, IN; San Francisco, CA; Boulder, CO; Aspen,
CO; and Nashville, TN all have high concentrations of degreed individuals.
That's pretty unsurprising. However, it is worth noting that those counties are small. What is more interesting is the western states with very large
counties. For example, you can see in Arizona that Maricopa and Coconino
counties have a slightly above average concentration of individuals with
bachelor's degrees. But the space on the map does not represent individuals. And according to the US Census Bureau, Coconino County only has a population of 134,500. These is important information we are missing because of the way the geographer chose to apportion the map and the information not captured or communicated.
In much the same way that people
began identifying themselves using statistical terminology, the problem with
geodemographics is that the same thing will likely happen because of how products are
marketed and information is distributed. Part of the danger of this form of
data gathering and marketing is that individuals will begin to view their needs
and desires based on the geodemographic system information available to them
(e.g. family planning decisions).
No comments:
Post a Comment
Note: Only a member of this blog may post a comment.