What do People prefer and What is more effective for Maps: a Decision making Test - Advances in Cartography and GIScience - page 168

Data used in the test is from the “Pastoral da Criança 1 ” database 2 , cor-
responding to real data about three different health indicators, which are
assumed to be for the years of 1995, 2000 and 2009. These indicators are:
“malnourished pregnant woman”; “child mortality”, and “malnourished
recent-born children”. For the choropleth and proportional symbol tech-
niques, data was standardized, while for dot maps absolute values were
employed. All data was collected at parish level, which is the smallest area
unit for this database. In order to make possible the comparison of the
three mapping techniques, the size of area units was made uniform, as not
to do so would possibly have had influence on the user's reasoning on
phenomena analysis.
Nine choropleth maps were built and three maps per theme were organized
in an A4 sheet, each one referring to a different year ( Figure 1 ). In an ana-
logue way, nine range-graded proportional symbol maps ( Figure 2 ) were
also disposed in three A4 sheets, each page related to a different health
indicator. For dot maps ( Figure 3 ), six maps were built, and the arrangement
was to dispose three different health indicators per A4 sheet, for two years:
1995 and 2009. Data was classified using optimal Jenk's method for four
or five classes, depending on the indicator and year. These maps are
intended to support the decision-making of the subjects and all of them
used the same data and stimulated the same mental information-processing
activities.
The universe of this experimental research is the universe of graduation
students of Paraná - Brazil. Considering a simple random sample, deter-
mined using 50% as the true population proportion, a questionnaire was
administered to 269 students in Curitiba, Brazil. Groups of between 13 and
40 students were tested at the same time, in similar classroom environ-
ments. It was assured that these classrooms were well-lit and with tables
sufficiently large to allow the use of four A4 sheets on each one.
Users were randomly assigned to one of the three experimental groups:
group 1 consists of subjects making decisions based on the nine choropleth
maps; group 2 consists of subjects using the nine proportional symbol
maps; and group 3 consists of subjects assigned to use the six dot maps.
Groups answered the same questionnaire, which was designed in order to
simplify the process of measuring map effectiveness, using the task scores
as this measurement.
1 Pastoral da Criança is a non-profit organization related to the Catholic church concerning
community health care, especially for underprivileged families and children.
2 www.pastoraldacrianca.org