Commonly, empirical evaluations of complex decision support systems are either suggested as deliverables of future work or limited to the self-reported satisfaction of the systems' users. Although self-reported satisfaction provides one perspective from which to evaluate decision-support systems, it also introduces a conflation of the user's satisfaction related to the decision support system on the one and the decision making process and its outcomes on the other hand.

This research explores alternative approaches to the empirical evaluation of decision support systems. The dissertation has two primary objectives. The first is to determine novel analytical methods for the empirical evaluation of web-based decision support systems based on human-computer interaction. The second is to contextualize individual decision making behavior within online transportation planning along the dimensions of egoistic and altruistic decision making.

The two objectives are operationalized in three analyses of a web-based decision support system for public participation in transportation improvement programming. In the first analysis, client-server interaction using multiple sequence alignment and hierarchical cluster analysis is evaluated. The findings suggest reliability of the hierarchical cluster analysis classification in regard to overall duration of interaction. In the second analysis, the statistical association of individual level variables and clusters of overall duration of interaction is investigated using logistic regression analysis. The regression results indicate a significant statistical association between overall duration of interaction and prior experience with online transportation discussions. Finally, in the third analysis, the extent to which egoistic and altruistic decision making expresses itself in the transportation project choices of the users of the system is analyzed. The choices are analyzed by examining locations of transportation projects selected by the users and their associated implementation costs in relation to the home, work, and travel locations of the users. The findings evidence user preferences for transportation projects within the vicinity of the individual's home, work, and travel locations as well as cost-aversion behavior in which the majority share of project implementation costs is carried by the average resident rather than the user. These observed behavioral patterns suggest predominately egoistic choice making behavior of the analyzed individuals.