In a two-way anova with interaction, we consider multiple comparison tests for the differences among main effects．We may derive a single-step test procedure based on Tukey (1953). Under simple ordered restrictions of main effects, we may also derive a single-step test procedure based on Hayter (1990). We propose closed testing procedures which are superior to the single-step procedures.

In this paper, I analyze the actual conditions of rest time from the end of work to the start of the next day from the relationship with the attributes of workers, etc. from the micro data of Survey on Time use and Leisure Activities. And analyze the influence of the rest time on the health status. It is the first attempt to analyze the effects of the rest time on the health from the data of statistical surveys. As a result, in addition to gender, it was found that the working time arrangement, educational background, and occupation are connected with the length of the rest time. And the shorter the rest time, the higher the probability that health condition will deteriorate, and commuting time it was found that the tendency is remarkable when it is added. And, the age, the number of days of paid holiday, and the delay of the start time of the rest time also affect the health condition.

This paper introduces the Spatial Computable General Equilibrium (SCGE) model for estimating the economic impacts of natural disaters. The SCGE model has a potential to describe the supply-chain damages between industrial sectors and regions as well as the change in final demands. However, few of the previous researches focus on the applicability of the SCGE model in the disaster impact analysis. The methodlogical framework is required to appropriately utilize the data obtained from rare damage records and reflect them into the basic settings of the SCGE model. For this purpose, it is necessary to estimate production capacity losses in the very detailed spatial scale using the economic census data, as an input of the model, and to investigate the closure rules and the values of parameters on the SCGE model. We have establiesd the SCGE model for the period of a few months after the Great East Japan Earthquake, which corresponds to the period that supply-chain damages were enourmous, but the model still needs to be refined, such as in the area of improving the values of substitution parameters between the goods and services from different regions. This paper briefly summarizes the outline of the SCGE model development for disaster impact analysis and reports the updates of the values of substituiton parameters which significantly improve the model performance.

In Japan, household size is shrinking. Since there is a correlation between household size and housing size, progress in household size reduction can be a factor causing a mismatch with existing housing stock. It is required to understand the relationship between single-person households and housing size, but it is important to grasp the current situation according to region by statistical data because the distribution of single-person households and housing by size is regional. On the other hand, when conducting regional analysis using statistical data, data tabulated for each prefecture or municipality is often used, such as the results of the population census by municipality. In analyzing such regional data, it is necessary to pay attention to the “modifiable area unit problem” in which the analysis results differ depending on the size of the regional division and the area to be analyzed. In this paper, we calculate the correlation coefficient of an arbitrary area by moving window method using grid square statistics on the population census and GIS, and evaluate the spatial scale of regional analysis. Furthermore, based on this evaluation, we map the distribution of the relationship between single-person households by age and the total area of housing, and consider the regionality of the relationship between them.

It is quite important in business marketing to understand the geographical distribution of retail stores and its trend across time. As a basic study of GIS in retail location analysis, we propose a new method for measuring the degree of spatial concentration of stores. The new method, which is developed from the Gini coefficient and the Lorenz curve, enables us to conduct a statistical test of whether or not stores are globally and locally concentrated in a region. The method is applied to the distribution of stores in Shibuya ward using the multi-temporal store location obtained from telephone books. The result indicates that the proposed method is effective and will be useful for location marketing.

In this paper, we deal with a mdoel selection in multivariate linear regression models, based on minimization of the model selection criterion (MSC) that includes the generalized information criterion (GIC) and the generalized C_p (GC_p) criterion as a special case, when the dimension of the response variables vector may be large but still smaller than the sample size. Recently, the consistent MSC evaluated from the large-sample and high-dimensional asymptotic theory was proposed. The consistency of the MSC proposed in the previous study can be achieved whenever the dimension of the response variables vector is fixed or goes to infinity. Unfortunately, the consistency property was showed under the assumption that the true distribution of response variables is the multivariate normal distribution. In this paper, we show that the MSC proposed in the previous study has consistency under the violation of normality of the true distribution.