Abstract
The purpose of this paper is to explore the choice of the housing size. Housing size, on the supply side as well as demand side, is one of the most important factors. Particularly, we can not overexaggerate the main policy of the supply side depends on how to allocate housing size ratio of total housing stock to construct. In this study, housing size is classified into 4 categories in accordance with substantial residential size in real market. The ordered probit model is, thus, employed as the empirical analysis for this research. The major findings are as follows; all of the variables to denote household characteristics were more likely to be associated with the choice of housing size. However, among the housing unit characteristics, some variables(residential type, the previous housing unit residential period) were found to be significant predictor in determining housing size. The main concern of this study also is to examine whether the community locations are substantially meaningful to the choice of housing size. The findings confirm that the variable to represent the central community over the eastern is not statistically significant, but the variable to represent the western over the eastern is found to negatively significant predictor in the choice of housing size. This suggests that those who live in the eastern community in comparison with the western are likely to prefer even bigger size of housing selection.