Predicting University Students’ Public Transport Preferences for Sustainability Improvement

Abstract

We present our study on the subject of improving the sustainability of daily life concerning university students based on a representative sample from a rural college town in Hokkaido, Japan. As a method of collecting the data, we surveyed two hundred and fifty students with a set of questions about their transport preferences and habits. The interpretation of data exposed a significant statistical hypothesis that the students would use public transport more often if made available in certain hours. The increase in the rate of people in favor of bus services grew from 59.6% ($MOE_{95}=\pm 6.0$) to 70.8% ($MOE_{95}=\pm 5.6$) if asked indirectly. Separately, to provide an insight to decision-makers of the regional development, we performed a Bernoulli trial and then fitted a logistic regression to classifying records of the data-set. Both approaches analogously reached around 78% of accuracy in predicting students’ public transport preferences by a small set of questions.