Identifying User Preferences Towards Proposed Colombo Light Rail Transit Using Hybrid Conjoint Analysis Model

Abstract

Preferences of car-based commuters to the Colombo Metropolitan area towards the proposed Light Rail Transit (LRT) line was evaluated using two-stage conjoint analysis. Attributes related to LRT were prioritized from an expert survey and categorized into levels. An orthogonal matrix was employed to develop 20 LRT profiles. User preferences towards the LRT profiles were modeled with an additive attribute only conjoint model. ‘E-services’ attribute was found to be the most important factor with 31.45% followed by the ‘accident safe’ attribute with 21.45%. Emergency evacuation system (+17.278), Online seat booking facility (+16.661), Reclining Seats (+15.051) were found to be the most important attribute levels which increased the utility of LRT while onboard WIFI facility was most disliked attribute level. Residual of the model were aggregated at user-level and regressed again using a linear regression model. Males and private vehicle owners were found to have significantly lower utility values 19.945 and 31.895 respectively.