Disaggregate mode choice models normally take account of transfers as the number of times the act will have to be carried out for a specific alternative. It was hypothesized that the prevailing transfer conditions and the use and non-use of transfer facilities characterize the quality of transfers, which in turn can have substantial effects on travel behavior. This research thus investigated the effect of the inclusion of possible measures of transfer inconvenience on disaggregate mode choice models for work trips. This was operationalized through the consideration of the different aspects of transfer such as walking, waiting, security risks, exposure to the elements and the use of stairs, with the latter forming an intergral part of the rail based option. Models with and without the transfer variables were then estimated using combined revealed preference and stated preference data. It was found that adopting the variables for the use of the stairs, maximum tolerable walking and waiting times and a generalized attribute for transfer inconvenience improve the specification and add significantly to the explanatory power of the mode-choice models.
