Some applications of travel demand forecasting require a longitudinal perspective of multiple days. The decisive advantage of the longitudinal perspective is the ability to analyze stability or variability of travel behavior. For measurement of stability, adequate indicators are required. A set of indicators for measuring stability of mode choice behavior was developed. A set of indicators is necessary because stability of mode choice has more than one facet, namely, frequency, change, and repetition. The indicators used so far to measure stability or variability of mode choice behavior consider only one aspect of stability, the frequency of modes used. Therefore, other indicators intended to capture the other aspects of stability are proposed. Applying the proposed indicators to synthetic mode choice sequences shows that they also measure aspects of stability other than frequency. For practical reasons, the set of indicators used to measure stability should be small. Correlation analysis reveals that the frequency-based stability indicators are highly correlated; therefore, it is sufficient to include one of the frequency-based indicators in the indicator set. The resulting indicator set allows the assessment of the stability of mode choice behavior for different notions of stability. As such, it might serve as a valuable tool for the calibration and validation of multiday travel demand models.
Measuring Stability of Mode Choice Behavior
Transportation Research Record
2017-01-01
Article (Journal)
Electronic Resource
English
Tourism Traffic Mode Choice Behavior Model
ASCE | 2011
|Analysis of hurricane evacuee mode choice behavior
Online Contents | 2014
|Analysis of hurricane evacuee mode choice behavior
Elsevier | 2014
|Choice Set Formation Behavior: Joint Mode and Route Choice Selection Model
Transportation Research Record | 2016
|