Estimation of average place attachment level in a region of Japan

In this study, we aim to develop a statistical model (place attachment model) for estimating the place attachment level for each of 241 cities by using available census data in Japan. This model can easily estimate place attachment levels for various regions where census data have been published. To develop the place attachment model, we used survey data in the Kanto region of Japan (N = 678). In testing the place attachment model we developed, we found several significant factors that contributed to predicting place attachment (i.e. age, length of residence, household income, marital status, land ownership, number in household, length of road per area, and population). Using this model, we estimated the average place attachment level in 241 cities. Additionally, the place attachment level was visualized on a map using a geographic information system. We found similar place attachment levels to be statistically clustered in neighboring regions.