SEEING THROUGH SMALL MULTIPLE MAP DISPLAYS
S.I. Fabrikant1, D.R. Montello2, S. Rebich2
1 - University of Zurich, Department of Geography, Zurich, Switzerland
2 - University of California Santa Barbara, Department of Geography, Santa Barbara, U.S.A.
As real-time three-dimensional landscape fly-throughs and interactive map animations of various spatial diffusion processes become ubiquitous with the dissemination through the Internet, one important question that remains is how effective the potential increase of information density in these highly interactive visual forms really is for knowledge construction and decision-making.
We still know very little about how effective novel interactive graphical data depictions and geovisualization tools are for knowledge discovery, learning, and sense making of dynamic, multidimensional processes.
A recent review of the cognitive literature on animated graphics suggests that animations are not superior to a series of static displays for conveying complex dynamic processes (Bétrancourt et al., 2000; Bétrancourt and Tversky, 2000; Morrison et al., 2000; Morrison and Tversky, 2001).
These cognitive scientists argue that experimental studies reporting advantages of animation over static displays lacked equivalence between animated and static graphics in content or experimental procedures.
Utilizing the eye-movement data collection method to track people’s viewing behavior, we investigate whether static small-multiple map displays are indeed equivalent in information content compared to non-interactive animated maps, as claimed by above cognitive scientists. If this were true, then people’s gazes on small-multiple displays would have to move sequentially from one map to the next in the display, matching the sequential viewing order users are locked into when viewing non-interactive animations.
In an ongoing experiment we first ask novice participants to study a series of small multiple maps showing monthly ice cream consumption for an average year for different states in a fictitious country, and then answer a number of questions about these maps. The test questions (within-subject independent variable) vary in type (e.g., space, theme, and time) and complexity (individual value retrieval, spatial inference tasks, including pattern detection and trend prediction over space and time). For example, participants are asked to identify the year (time) or the location (space) of maximum/minimum ice cream consumption (theme), the maximum/minimum change in consumption (time), or describe a potential consumption pattern, and predict a possible consumption trend. We prompt viewers to think-aloud while performing these spatial inference tasks. Digital audio recordings of their verbal statements while answering the test questions permit joint analyses with accuracy of response, and participants’ eye movement recordings (dependent variables).
With the collected empirical evidence we hope to provide better understanding of how people use static small multiple displays to explore dynamic geographic phenomena, and how people make inferences from static visualizations of dynamic processes for knowledge construction in a geographical context.