How does accessibility predict trip production?
Sylvin, Joni (2024)
2024
Matematiikan ja tilastollisen data-analyysin maisteriohjelma - Master's Programme in Mathematics and Statistical Data Analytics
Informaatioteknologian ja viestinnän tiedekunta - Faculty of Information Technology and Communication Sciences
This publication is copyrighted. You may download, display and print it for Your own personal use. Commercial use is prohibited.
Hyväksymispäivämäärä
2024-11-29
Julkaisun pysyvä osoite on
https://urn.fi/URN:NBN:fi:tuni-2024112510477
https://urn.fi/URN:NBN:fi:tuni-2024112510477
Tiivistelmä
The aim of this work is to study the effect of accessibility as well as other explanatory factors on trip production with statistical modelling. Previous zonal and and national traffic models used in Finland have been found to be lacking for which reason the Finnish Transport and Communications agency Traficom has been assigned to build a national traffic modelling system which this work is part of. Transport models are typically divided to four levels including trip generation, distribution, modal split and assignment with distribution and modal split being often together nowadays. Trip production refers to number of trips made by person during a day or a year.
The data for this study are a survey of a three week period of long distance leisure trips of distance over 100 kilometres with the explanatory variables in data including e.g age, income level, car ownership and employment status. Location information is also included in form of zone id and areal classification variable.
Accessibility measures how people value different opportunities available to them in their action zone. The accessibility indicator for this work is derived from random utility theory and multinomial logit model. Concerning accessibility, the area of leisure apartments, number of jobs in hotel and restaurant industry and population are taken into account in the measure as well as price and time of trip.
The methods used compare different statistical models of which zero-inflated and hurdle model variations proved to be the most fitting ones for this study. The results indicated that the used accessibility indicator does not have a statistically significant effect on trip production in any of the tested statistical models. The limit of 100 km for trips already in trip production stage is limiting the direction of trips and might therefore result in distortions in models. It was also found that income level, car ownership and employment status have a significant role in trip production levels. These findings would indicate that in addition to increased economical resources, available time resources also contribute increasingly to the number of long distance leisure trips made by an individual. For future studies, it could be considered to change the limit of trip length or for instance study a different trip category.
The data for this study are a survey of a three week period of long distance leisure trips of distance over 100 kilometres with the explanatory variables in data including e.g age, income level, car ownership and employment status. Location information is also included in form of zone id and areal classification variable.
Accessibility measures how people value different opportunities available to them in their action zone. The accessibility indicator for this work is derived from random utility theory and multinomial logit model. Concerning accessibility, the area of leisure apartments, number of jobs in hotel and restaurant industry and population are taken into account in the measure as well as price and time of trip.
The methods used compare different statistical models of which zero-inflated and hurdle model variations proved to be the most fitting ones for this study. The results indicated that the used accessibility indicator does not have a statistically significant effect on trip production in any of the tested statistical models. The limit of 100 km for trips already in trip production stage is limiting the direction of trips and might therefore result in distortions in models. It was also found that income level, car ownership and employment status have a significant role in trip production levels. These findings would indicate that in addition to increased economical resources, available time resources also contribute increasingly to the number of long distance leisure trips made by an individual. For future studies, it could be considered to change the limit of trip length or for instance study a different trip category.