Description
This research compared the effectiveness of different colors and color schemes for 3D urban spatio-temporal map animations. The animations have been created using ArcScene and Adobe Flash, each re-illustrating urban growth of San Diego State University campus buildings from the 1950s through the 2000s, where campus buildings varied in hue or value for each subsequent decade. The three main animations that have been statistically compared consisted of a red sequential scheme, a blue sequential scheme and a random hue color scheme. The following three questions have guided the framework of this study: (1) How do different colors and schemes (i.e. red versus blue and sequential versus random hue) compare in affecting user abilities to acquire spatio-temporal information in 3D urban map animations? (2) What other factors (i.e. exaggeration and a Google base map) affect user abilities in acquiring spatio-temporal information in 3D urban map animations? (3) How do findings from eye-tracking analysis reinforce or contradict observed results of specific color schemes? Participants viewed one of the three randomly assigned animations and completed a flash based survey testing their ability to recognize and comprehend spatio-temporal information presented in the 3D map animation. In total, 94 users were tested, with 35 for the blue color scheme animation, 33 for the red color scheme animation, and 35 for the random color scheme animation. Statistical results reveal no statistically significant differences between the overall mean accuracy survey scores when comparing the red and blue sequential color scheme animations, nor the sequential and random hue color scheme animations. It should be noted, however, that participants viewing the red animation survey performed the worst overall (i.e. survey accuracy), suggesting a possible issue with the low sample sizes and its connection to the lack of statistical significance. Upon comparing individual questions for each survey, participants viewing the blue animation outperformed the participants viewing the red animation for two of the survey questions with a statistically significant difference between the mean scores at the .10 level of significance. Additionally, two further questions yielded statistically significant differences (p-value = .05) upon comparing the sequential and random hue animations, where the random hue performed the best in one question, and the sequential scheme in another. In the final portion of the survey, participants provided additional qualitative feedback regarding three further animations that utilized the same color scheme in which they had been randomly assigned to. The first animation exaggerated building heights by two times their actual height in scale, another by four times, and a final animation utilizing a Google base map footprint. Overall, the majority of users preferred the animations that exaggerated building heights and disliked the animation employing the Google base map. Lastly, with regards to eye-tracking analysis, nine users were tested, three per main animation. These results indicated a necessity in additional testing as the findings mostly indicate individual differences (rather than between animations) among each user. Eye-tracking analyses indicate a great potential for future research in explaining the survey results. Additional exploratory research has resulted in several notable findings: 1. Users who are able to watch the animation multiple times were able deduce more spatio-temporal information than watching the animation only one time (as indicated in their accuracy scores). 2. User answer response times and accuracy scores appeared to have a positive correlation, where as the answer response times increased, so did the accuracy scores. 3. Male users outperformed female users based upon the statistically significant differences (p-value = .05) in their overall mean score. While very few significant results were found, several important research questions have emerged, opening new possibilities in future 3D map animation research.
