Cognitive Issues in Geographic Information Visualization

Perception and Thematic Map Symbol Design II

4108.3 - Cartograms in the Making: An Experiment in Aesthetic Design

Presenting Author(s)

• BK

  Barry Kronenfeld

  Associate Professor
  Eastern Illinois University

Co-Presenter(s)

• DV

  David Viertel

  Associate Professor
  Eastern Illinois University

Moderator(s)

• 

  Claudia Sluter

  Professor
  Federal University of Parana - UFPR

Motivation

Understanding the cognitive properties of cartograms and improving their appearance such that they produce less cognitive dissonance remains a primary challenge in cartogram research. A few recent studies have shown that favorability of human subjects differs depending on the form of the cartogram, with some preference for cartograms with rectangular form. However, these studies are limited in that they compare only a small number of cartograms among infinitely many possibilities.

Since cartograms rely on recognition of shapes that may be morphed in an infinite variety of ways, a broad range of forms must be explored to discover characteristics of cognitively salient and aesthetically pleasing cartograms. It is difficult to produce such variety using existing algorithms. We take a different approach, providing subjects with tools to create their own cartogram and then examining the choices that they make.

Research Questions

Given complete control over cartogram form, we posit that subjects will focus more attention on features considered cognitively salient and to exhibit aversions to types of distortion that create greater cognitive dissonance. Based on these concepts, we develop several specific hypotheses that will be tested experimentally, namely that subjects will focus more on:

(1) "linear" boundaries than non-linear boundaries
(2) external boundaries than internal boundaries
(3) shape "parts" than proportionality

These are working hypothesis only; our experimental design leaves room for the possibility of other discoveries that cannot be anticipated in advance.

Methods

Seven subjects in an upper-level undergraduate course in cartography will participate in a class assignment to construct a cartogram of the U.S. states based on Electoral College representation. Cartogram construction will be performed in Cartogram Studio, a flexible mesh-based software application that provides tools for interactively constructing a continuous cartogram by moving and adjusting map features. After presenting historical background, uses and methods of cartogram construction, students will first spend one 2-hour class period to learn to use the software. In the next 2-hour class period, students will be asked to construct a cartogram of Electoral College representation of the conterminous U.S.A. (48 states and the District of Columbia).

Subjects' focii of attention will be assessed in two ways. First, following the assignment students will be asked in a class discussion (or written assignment?) to identify specific locations that they focused on during cartogram construction. Second, individual adjustments made by each student will be examined to identify "hot-spots" of attention, on the assumption that a greater number of adjustments indicates a greater degree of cognitive focus. Students will be aware that their actions are being recorded, but the analysis of these actions will occur independently after the end of the semester and will not affect their grades.

Anticipated Results

The results of this study will improve our understanding of the relative cognitive salience of different aspects of shape and shape distortion on cartograms. This may help to improve cartogram construction algorithms by (1) identifying priority areas for shape preservation, and (2) informing the definition of appropriate shape distortion metrics. Given the small sample size, any conclusions will likely be preliminary. On the other hand, the small sample size allows us elicit a much greater degree of participation from each participant and to conduct an open-ended "focus group" discussion to elicit further insight. We hope that the insights gained from this experiment can be used to develop briefer test/survey instruments that will allow us to employ a larger subject pool in future experiments.