Research On The Car-hailing Travel Behavior Of Elderly People Based On Latent-class Conditional Logit Model

With the increasing average life expectancy of human beings,the population is gradually showing an aging trend all over the world.China,where the birth rate is decreasing and the number of elderly people is increasing,is no exception.Meanwhile,thanks to the innovation of information and communication technology,the rapid development of transportation modernization in China has brought people digital and intelligent travel modes that provide more convenient,effective and comfortable travel services.For example,car-hailing is becoming more and more popular because it can meet the demand of peer-to-peer travel.The“Digital Divide”,however,may keep the elderly away from the quality services brought by modern transportation development unequally.Compared with young people,they probably choose other inconvenient transports due to lack of the ability of accepting and applying new techniques.Therefore,more attention should be shown on their travel demands and behavior.Travel behavior can be analyzed through the study of travel mode choice.Most of the current studies use multinomial logit models that tend to ignore individual preferences.Some studies classify samples based on potential heterogeneity,but the classification is mostly done by demographic characteristics.However,Latent-class Conditional Logit Model allows for objective classification based on individual preferences.It assumes that samples can be classified into multiple latent categories,with individuals in the same category having the same preference parameters,rather than an artificially determined categorization.Therefore,this research focused on the car-hailing travel behavior of the old people to explore the distinct affecting factors.The data of relevant personal and travel selection information for the old,which was collected randomly through the questionnaire in Changsha city,Guangzhou city and Jilin city,China,was fitted by using “Stata” and “R Language” programs based on Latent-class Conditional Logit Model.It was found that although the aged tend to taxi rather than car-hailing in general,especially for those aged 65 to 74,the willing to select car-hailing would be improved significantly while they own better education level,smartphone and higher phone proficiency.With the further research,Latent-class Conditional Logit Model showed improved goodness of fit and accuracy of prediction.Its fitting results displayed that the aged could be classified into three latent classes,namely,class of car-hailing-propensity,class of taxi-propensity and class of neutral-propensity.These three latent classes accounted for 18.7%,34.8% and 46.5%of the total interviewed subjects,respectively.Furthermore,those with college education or above,high phone proficiency and the ability to use electronic payment or positioning navigation function were more likely to be defined as the class of car-hailing-propensity.Meanwhile,those aged 65 or above,living alone,without high school experience,and without smartphone were probably to be considered as taxi-propensity class.Suffering from chronic diseases,not living alone,and poor economic situation would increase the probability of being the neutral-propensity class.Besides,ordered logit model was used to further validate that educational level,phone proficiency and phone functions provide the motivation for car-hailing.Moreover,a comparative analysis of the data fitting results for Changsha city,Guangzhou city and Jilin city was conducted in the last part of the research.In these three cities,where the senior citizens could be distinguished as the above latent classes respectively,the shares of the neutral-propensity were the highest while that of the car-hailing-propensity were the lowest.The factors that conspicuously affect the car-hailing-propensity were highly consistent as well as their influence degree.In addition to the said positive impact,age,gender,economics,cohabitation and smartphone ownership were less apparent or varied among cities.According to the comprehensive analysis results,seniors with higher education were more motivated to travel by car-hailing.Even though more than 70% of the seniors had smart phone right now,the proficiency of related functions was the key to remove the restriction on car-hailing.