Africa, parts of Asia and some of China, includ-
ing Hong Kong.
Many geographers have been active in
developing geodemographic classifications, in-
cluding Super Profiles (Charlton, Openshaw
and Wymer, 1985), GB Profiles and a freely
downloadable classification of UK Census
Output Areas (at http://neighbourhood.statis-
tics.gov.uk). Others have been more critical.
One concern is that for some applications the
cluster groups are not sufficiently homoge-
neous for them to represent well the individ-
uals (or households) allocated to them. Voas
and Williamson (2001) suggest that apparent
differences between geodemographic classes
conceal a much greater diversitywithinthe
classes. A related concern is that the montage
of variables forming a geodemographic classi-
fication creates something of a black box,
making it hard to determine the key predictors
of the geographical phenomena being ana-
lysed. Care needs to be taken when interpret-
ing geodemographic outputs because they are
usually indexed as rates in one cluster, relative
to all others. To find that an event is of above
average prevalence in one geodemographic
group is no guarantee that it is common
there: the result could apply to a small minor-
ity of the population but still a larger propor-
tion than for other clusters.
Surrounding geodemographics are broader
debates inhuman geography, including those
aboutrepresentation, quantitative method-
ologies, empiricism, generalization, induc-
tionversusdeduction, data- versus theory-
led approaches to understanding,neo-liberal
economies and the politics and commercial-
ization of data collection, privacy and social
discrimination. Critical theorists have cited
geodemographics as an example of ‘software
sorting’, suggesting that the sorts of labelling
used in geodemographic systems can produce
stigmatization of certain places and potentially
deny them the same level of (e.g. banking or
insurance) service given to otherneighbour-
hoods(Burrow, Ellison and Woods, 2005).
However, the argument cuts two ways: geode-
mographics can also identify areas of social or
material need, offering opportunity to better
target the resources available to those places.
Geodemographic classifications can be used
to interpolate market research and other sur-
vey data to standard administrative or ad hoc
geographies allowing, as examples, estimation
of: the levels of consumption of grocery prod-
ucts by supermarket catchment; demand for
particular makes of car by dealership territory;
or likely levels of diabetes by GP surgery
catchment area. Whereas much academic de-
bate centres on the accuracy (or otherwise) of
geodemographics for predicting the behaviour
of individuals, in practice many users are inter-
ested in aggregate behaviour – What, on aver-
age, is the most likely event, characteristic or
behaviour at an area level, and how does this
differ from other areas?
Increasingly, uses of geodemographics bring
together academic, public policy and private-
sector stakeholders, applying geographical
thinking to tackle questions of social concern.
Geodemographics has stimulated a renais-
sance in applied geographical research, being
recently used for investigating the spatial dis-
tributions of family names, predicting spatial
variation in pupils’ school examination per-
formances, examining inequalities in hospital
admissions and for guiding localpolicing(all
at http://www.spatial-literacy.org). rhSuggested reading
Charlton, Openshaw and Wymer (1985); Harris,
Sleight and Webber (2005).Geographic Information Science (GISc) In
the simplest sense, Geographic Information
Science (GISc, or GIScience) is the theory
that underliesgeographic information sys-
tems(gis). The latter are the collection of
hardware, software, output devices and prac-
tices are that used to analyse and map spatial
entities and their relationships. GIS software
might be used to determine theboundaries
that distinguish areas with different average
income levels in a city or a map of optimal
delivery routes for a courier company. These
results are, however, not transparent; the pro-
cess through which they are derived are known
asblack box. Geographic Information Science –
or the theoretical basis for GIS – is concerned
with how results are obtained in GIS and what
questions can legitimately be asked.
GIScience explores how spatial objects be-
come digital entities, what effect that trans-
formation has on their digital ontology,
how different epistemologies affect onto-
logicalrepesentation, how to model relation-
ships between spatial entities, and how to
visualize them so that human beings can inter-
pret the results (Raper, 1999). This pursuit
draws on and extends developments in data
modelling, computer science, cognition,visu-
alization and a myriad fields that have
emerged in response to information systems.
For the first several decades of GIS use,
little attention was given to the differentiation
between geographical information systemsGregory / The Dictionary of Human Geography 9781405132879_4_G Final Proof page 277 2.4.2009 6:30pmGEOGRAPHIC INFORMATION SCIENCE (GISC)