In Japan, Many Geographers have been attempted to make GIS programs such as Nogami (1990a), Ogata (1987), Yoshimoto (1990) and so on. Today it is important for us not only to use GIS as a mere mapping system but also to make use of it by integrating with analyiical system for geographical study and education. In this paper, we report one example of such integrated system on personal computer, PC-ARC/INFO + PC-SAS. Next, we analyse the regional structure of agricultural income of Kinki area using this system. <<STATISTICAL ANALYSIS + GIS>> First, factor analysis of 15 variables is executed, and factor scores are represented by GIS. Factor I (productivity and fruit agriculture) has high score in metropolitan core area. Factor 2 (grain agriculture vs. stock raising) has high score in plains and baisns of metropolitan peripheral area. <<GIS + SPATIAL ANALYSIS>> Second, GIS produces the adjacency of each district and spatial autocor-relations (Moran statistic Is) of 15 variables are calculated. The distributions of variables such as productivity and grain percentage have tendency of clustering, so Moran statistic Is of these variables have high values. The integration of GIS and analytical system will provide not only a tool of mapping but also a tool of our thinking about geographical problems, because it helps to connect the processes of mapping and analysis mutually in our studies.
In the Japanese Metropolitan regions, most of the housing supply comprises of both housing estates in the suburbs and residential buidings. For example condominiums and privately or publicly owned houses in built-up area. The housing supply has some sirnilar kinds of housing types in each unit. Consequently these urLits of the housing supply standardizes the characteristics of their dwellers, as to age of householders, family composition and the size of household of life cycle. Because the houses that are supplied are standardized in price, dwelling space and have a sirnilar place utility. This paper aims to examine the differentiation of the residents in the context of mixed housing supplies, and to clarify the characteristics of each resident in housing estates, condominiums, publicly owned houses, privately owned houses and in company's houses for employees in Fukuoka city, which has a variety of housing supply types. The results are summarized as follows; 1. According to the 1980 National Census, many middle aged householders live in their own houses in the suburban housing estates. In the older housing estates which were built before the 1970's, the elderly live in their own houses as opposed to the newly built ones, because the residents in housing estates have the tendency to live there all their lives. On the other hand, younger householders tend to rent a house in privately or publicly owned housing sectors, because they can not afford to have their own houses. Publicly owned houses near the inner city are occupied by many elderly householders, but in suburban publicly owned houses many younger householders live. 2. Larger size households tend to live in housing estates, company's houses and privately owned houses in the suburbs. Because the suburban houses have more space for the fainily than publicly owned houses and condominiums in the built-up areas. Though all of publicly owned houses are too small for large size households to reside in, Iarge size households are given priority to publicly owned houses, which must play as important role in welfare. 3. In the light of the labour force status of residents, the management tend to live in their own houses such as those in suburban housing estates or condomirLiurns in the built-up areas. Considering the residents in privately owned or companys' houses, the rates of physical labour force are high. This defferentiation is caused by economic status connected with labour force status. Regardless of the householder's age, the household size and the kinds of houses, service employees tend to reside within the iuner city. This is because service employees think much of accessibility to their working places. 4. A Iot of residents in housing estates and condominiums migrate from the neighbourhoods and stay 10nger in their own houses. On the other hand, residents in companys' and privately owend houses tend to migrate from other prefectures and stay shorter. Therefore, some of the differentiations of dwellers characteristics are recognized and some evidence obtained shows the relationship between housing tenure and the age of householders, the size of household, and the occupation of householders. These are caused by the policy of housing supply that different households are supplied with different housing types according to their characteristics discussed above. Such a selection causes homogeneity of housing units residents.