Logistic regression in modeling and assessment of transport services

Abstract

AbstractTransport companies operate in a dynamically developing and competitive market. Maintaining the already held position and further expansion requires adjusting the level of services provided to the needs and requirements of customers, as well as continuous surveying, monitoring and adjusting the implemented strategy. There are different methods for such an analysis. This article proposes logistic regression. The research was conducted on the basis of a distribution and trade company dealing with the supply of automotive spare parts. As the most profitable group of customers are local car repair shops, it was this group that was subject to analysis.The quality of service assessment was considered from the point of view of delivery time. The dichotomous form of the predictor taking two values - late and on-time delivery - was determined. From among the possible ones, regressors whose influence was statistically significant and whose modification was possible were selected. The research showed which of them (and how strongly) affect the dependent variable, which allowed for modification of strategy and implementation of new solutions increasing the number of satisfied customers.