Geography Reference
In-Depth Information
share of 70.1 % (share of 5 % or more for each individual topic): Population—
Territory and Environment—Industry and Services—Agriculture and Forestry—
Tourism—Politics—Historical Statistics (long-term times series).
Due to the use of own primary data sources (around 85 % of all data used in
official statistical maps come from the NSS) statistical maps and atlases reach a
degree of topicality unlike any other thematic atlas type. Around two third of the
maps are no older than 2 years, one tenth no older than 1 year (Schulz
2013
).
Cartographic Representations
Not without good reason, many authors use the term “choropleth map” synony-
mously with “statistical map” (cf. Hake et al.
2002
or Bill and Zehner
2001
). The
overwhelming availability of elementary analytical data in the public statistical
domain correlates unambiguously with the cartographic representation methods
applied within their products. There is a direct link between the 98 % simple key
figures used as data bases and the 98 % analytical representations that result in these
cartographic works (Schulz
2013
). Only three representational methods dominate
statistical maps and atlases today:
area cartograms
,
proportional symbol
cartograms
and
combinations
of both. Other methods (incl. “statistical methods”
as named by Spiess et al.
2010
) remain clearly at the margin—independent of the
regional or temporal context. Schulz (
2013
) analysed app. 6,300 maps in 20 statis-
tical atlases. In total, the above mentioned representation methods count for 89 % of
all maps found in these atlases. Another 5.5 % are qualitative maps. The statistical
atlases taken into account by this study come from regions all over the world and
from different historical periods, going even back to the nineteenth century. Thus,
the results can be generalised and are not only valid for Switzerland or Europe.
These results correspond also to other studies conducted by e.g. Freitag (
1988
), van
Elzakker et al. (
2005
), or Dickmann and Zehner (
2001
). Van Elzakker et al. (
2005
)
analysed statistical maps and atlases that were published on the Internet by NSIs
around the year 2002. The most common maps found on these websites were
choropleth maps and proportional symbol maps of all kinds.
The use of seemingly endless rows of simple choropleth maps in statistical
publications, often criticised and disliked by cartographic scientists and profes-
sionals, has structural and provable reasons linked to the data bases and user
demands. It can only to a very limited percentage be derived from the ignorance
of other semiotic means and methods by the respective authors and institutions.
Maps in statistical publications do primarily not serve scientific purposes, as
mentioned before. They are made for the layperson and the occasional visitor.
Their intentional simplicity in design and content makes statistical maps so suc-
cessful, as web statistics of several NSIs show. The Statistical Atlas of Switzerland,
