DATA NATURE AND EFFICIENCY OF CHOROPLETH MAPS
Chair of Cartography, Faculty of Geography and Regional Studies, University of Warsaw, Poland
Choropleth representation is one of the most widely used types of visualization in thematic maps. However cartographers still have not specified the principles of elaboration of choropleth maps to accurately represent a spatial pattern of a given dataset. One of the issues is the type of data to be used for choropleth map design.
Most authors researching choropleth methodology point out that the most correct approach is when choropleth maps are based on relative data applied to the entire area of enumeration units (for example: population density). For choropleth map design cartographers also partially accept the use of other related data: amounts related to a part of the area of enumeration units (e.g. share of wheat sown area in the total sown area), amounts related to the total population of a unit (e.g. share of population aged 15–59 in total population), amounts related to a part of the population of a unit (e.g. share of the unemployed with elementary education in the total number of the unemployed) and amounts related to other elements (e.g. number of inhabitants per pharmacy).
Cartographers have not agreed on the use of absolute data for choropleth map design. Some authors seem to intentionally avoid this topic, while others do not allow such option at all.
Validity of cartographers’ theses has been verified in an experiment performed at the Faculty of Geography and Regional Studies at University of Warsaw. The efficiency of choroplethic transmission of information, depending on data type, was evaluated by a test. Students were divided into four groups. Each group was presented with a set of choropleth maps and block diagrams based on different data. The efficiency of choropleth maps based on various data was tested on three levels of map reading: general, detailed and intermediate. Information about three dependent variables was collected: accuracy (measured as percentage correct), reaction time and confidence (of respondent that the given answer was correct). The results of experiment are presented in the paper.
The experiment helps to define diffuclties in transmitting information with choropleth maps. It also helps to evaluate efficiency of the choroplethic method of visualization in the process of information transmission.
The results of the experiment can be useful for the development of GIS software, which gives the cartographically untrained map user wide possibilities of symbol selection and at the same time reduces the risk of a potential misuse of cartographic methods of representation.