|Abstract: ||Recent developments in information and communication technology have meant that mobile commerce will have an increasingly important influence on the travel behaviour; in particular on how, when and where people choose to use their time to conduct activities. In the cases of activities involving the use of mobile services, such as m-shopping or m-banking, the conventional strong connection between the utility which one derives from an activity and its timing and duration are weakened or even broken such that existing utility-maximization models are not suitable in these contexts. This thesis presents a novel approach to characterising the utility of activities which can be applied in a broader set of circumstances, especially those such as m-services in which technology relaxes the patio-temporal constraints of activity participation.
Building on concepts from both the microeconomic literature and the activity scheduling literature, the thesis proposes a unified utility framework based on an activity production approach, which is characterised by an activity production function. In this approach, an activity is regarded as the archetypal "small firm‟ theorised by Becker (1965). An individual transforms the inputs (purchased market goods, time, and technology) into the outputs (consumption 'commodities) through some activity production process. Both the production process and the outcome of consumption are potential sources of (dis)utility. We show how this framework generalises existing activity utility model and demonstrate how it can be extended to accommodate activities performed in both electronic and mobile contexts.
A stated choice exercise was undertaken to obtain the data necessary to estimate utility models based on this new framework. In this exercise, a hypothetical shopping choice scenario was presented to respondents and a D-efficient design was adopted to investigate people‟s decision making with respect to shopping.
Using the cleaned data generated by this data collection program, the basic model with the best overall level of fit was identified. Theoretical and empirical implications were discussed. Strong evidences of taste heterogeneity among respondents were also found.|