summaries
J H. Geels - A model for selecting area symbols (part 2)
This is the second part of an article on a method for selecting
area symbols adhesive screens) in diagrams and maps. In the
first part some terminology was defined, in this part examples
of its use are given and a few rules are discussed. A summary
of the technical details is given below. Please refer to part one
of the article (in Kartografisch Tijdschrift I987.XIII.4) for a
more general summary. More information in English is available
from the author.
The design model discussed is graphically summarized in the
diagram.
measurement i
niveau de
lecture/
representation
level of
formation
ordinal
niveau de
mensuration
Visual variables
variables visuelles
the plane
detinition
definition
j implied
I„J impliquee
optional
optionnelle
texture
grain
couleur
Level of Measurement
Definitions:
Nominal scale: a set of identifiable objects without assumptions
about their relationships. Typical Operation: equivalence. Exam-
ple: religion.
Ordinal scale: a nominal scale with an order added. Typical
Operations: higher. lower. Example: socio-economic Status.
Interval scale: ordinal scale with a measure of distance added.
Typical Operations: add, subtract. Example: temperature.
Ratio scale: interval scale with a measure of proportion added.
Typical Operations: multiply, divide. Example: time.
Rules:
Try to use a pure Visual variable with a level of measurement
equal to the level of measurement of the data. If you choose a
Visual variable with a lower level of measurement, information
will be lost. If you choose a Visual variable with a higher level
of measurement, new information musl be created.
Discussion:
The levels of measurement form a hierarchy. This means that it
is always legitimate to drop back one or more levels. However,
you lose information in doing so. Particularly, information given
by the Operation typical of the level you abandon. Therefore it is
It
recommended to make use of the highest level of measurement
that can legitimately be assumed.
Constructing an image implies using size or value, but these are
restricted to the top levels of measurement: interval and ratio.
How to achieve an image for data on the bottom levels: nominal
or ordinal? Inflate the level of measurement yourselfl Nominal
becomes ordinal, when you adopt order for the data. Ordinal
becomes interval, when you adopt distance for the data. When
the data are on interval level value can be used to construct the
image. Essentially you add new information yourself in doing
so. This is a political decision.
Level of Formation
Definitions:
Identity: identifying a Single instance as belonging to a Single
class.
Selection: isolating all instances of a Single class.
Image: simultaneous perception of all instances of alt classes.
is the product of size times value. It emerges as an imaginary
hilly landscape, the Z-dimension.
Rules:
Improve the identity of a class by combining differential Visual
variables texture - colour - orientation - shape. Improve the
selection of a class by improving its identity to the ultimate. Im
prove the image by adapting order and distance to the distribu-
tion of the data.
Discussion:
The levels of formation do not form a hierarchy. Whereas selec
tion implies identity, image does not imply neither.
On the highest level of formation image is essential, selection
and identity are optional but recommended. When supplemen-
ting identity and selection to an image do not alter size or value
for they shape the image. Use differential variables only, as they
do not interfere with the image - except large variations in tex
ture. To remedy this, look at the graphic with pinched eyes. Tex
ture fades away and image reappears.
M.
Bergmeijer - Marketing in cartography, an exploration.
The importance of marketing theory for cartography is based on
the idea that map design, map production, map sales and map
use are mutually related.
In the field of cartographic research too little attention is paid
to the benefit of marketing know-how.
In the Netherlands marketing research has been recently carried
out by the sales division of the State Publishing House (Staats-
uitgeverij) at the onset of the production of the new national
Atlas of the Netherlands. This research was a good example of
the application of general marketing theories in a non-
traditional field as cartography.
Marketing activities in Canada, Great Britain and Australia also
have indicated that valuable information is to be gained on
buyer's attitude towards a cartographic product. It is shown that
relatively simple marketing activities have a positive influence
on map sales and the increase of understanding and acceptation
of cartographic products by the general public.
66