PLS path model building: a multivariate approach to land price studies: a case study in Beijing

Previous studies have revealed that statistical methods can be used to analyze land-leasing parcel data. However, the conventional statistical methods used in land analysis have some limitations, especially in cases of limited observational data. In this paper, with the help of geographic information system (GIS) techniques, a partial least squares (PLS) path model is applied to study the relationship between residential land prices and various determinants through a case study of Beijing in China. From a preliminary analysis, four latent variables are selected: accessibility of the workplace center, livability, traffic, and environment facilities. The results show that the observation variables have a strong explanatory power for their corresponding latent variables, and the four latent variables have varying impacts on residential land prices. Of the latent variables, accessibility to the workplace center has the strongest impact on the residential land price