This paper examines the effects of conservative political ideologies on attitudes to sciences and humanities in Japan. As the conservativeness is multidimensional, we measure it with four scales: self identification with liberal/conservative, support for a coservative regime (Abe cabinet), authoritarian attitudes, and chauvinism. The attitudes to sciences and humanities are measured with six scales: recognition of utility coming from sciences and humanities and relative trust on ecologists, life scientists, economists, historians, and scholars of constitutional law. The results show the tendencies that the more conservative are, the more negative to sciences and humanities, while several exceptions are observed. Conservative recognize more utility of sciences and humanities than liberal, and conservative trust life scientists as much as liberal.

This study investigates trends of educational assortative mating and its causes in Japan, where there has been a rapid increase in the unmarried population. Past studies, which have examined assortative mating as a measure of social openness, did not consider the influence of the rise in this population. Against these studies, I examined trends in educational assortative mating focusing on the role of marriage decline. A survival analysis using SSM2015 data reveals two points. First, a retreat from marriage has continuously increased among all groups, while there has been a decline in the marriage rate at age 40 among female university graduates in a recent cohort. Second, decomposing hazard rate of marriage into spouse’s educational attainment shows a relative change in couple’s educational coupling. Due to the women’s improved access to higher education that accompanies a relative decline in highly educated men, university educated women are more likely marry down in a recent cohort. Junior/high school and junior college graduates are more likely to marry a man graduated from junior colleges and less likely to marry am university educated man, while a marriage with junior/high school educated men did not increase. The results suggest that, through the persistent male breadwinner model and recent employment instability, educational assortative mating among the low and high educated women decreased in parallel with the trends in marriage delay and foregone.

Being self-employment decreasing in Japan, there is not much research on the withdrawal from self-employment. The hypotheses are set up by taking into consideration patriarchal ideology, the division of labor between men and women, and the abolition of the Large-scale Retail Stores Law (2000). The data used is the 2015 SSM survey data (the national survey of Social Stratification and social Mobility of 2015). The analyses focus on male and female self-employers. Discrete logit models are applied to “withdrawal from self-employment” for male and female, “closing a business” for male, and “withdrawal for family reasons” for female. The findings for male self-employers are that they are likely to close their businesses in old age because their sons do not take over, they are not likely to close if children live together, and they are likely to close after divorce. The findings for female self-employers are that they are not likely to withdraw after marriage and their spouses’ deaths, and they are likely to withdraw for childrearing and nursing. The Large-scale Retail Stores Law has no effect on the withdrawal. The decrease of self-employment is not related to the weakening of regulation, but related to the weakening of family-based social capital.

The application of statistical network analysis, such as Exponential Random Graph Models (ERGM) and Simulation Investigation for Empirical Network Analysis (Siena), has been widespread in social network analysis. In this article, the characteristics of these methods were reviewed, following which their usefulness was discussed. The advantages of statistical network analysis methods are multivariate modeling of the effects of the features of nodes or node pairs in networks upon the connections of nodes and the statistical significance tests of these effects. The recent development of statistical packages for these methods made it easy to apply them to real data; however, some research suggested that the usefulness of these methods is restrictive for the purpose of social network analysis, which investigates social structures. This article demonstrated that statistical network analysis methods mainly focus on the micro features of networks, which makes it difficult to model the macro features of networks, and that these methods assume that the effects upon the node pairs are uniform. Finally, the application of these methods to the analysis of network processes such as formation, development, and declination was proposed.

This study explores a context of competition between social venues and between health organizations that are structured by different patterns of client referrals. Using the economic concept of properties of services/goods (“substitutes” or “complements”) and the network concept of structural equivalence, we derived several hypotheses regarding the associations between specific referral patterns and the presence of competition ties. Referral ties may exist among social venues and health organizations to better serve their clients, such that social venues may help clients to identify other venues that are more central to the gay community and that health organizations may help clients to obtain services that they themselves do not offer. Based on the data of 20 social venues and 20 health organizations in Chicago, IL, and Houston, TX, we tested our hypotheses using exponential random graph models. The results indicated a co-occurrence of referral and competition relationships between social venues. This tendency, however, was not found between health organizations. Health organizations in a similar network context in terms of referral inflow from other health organizations tended to have a competition relationship. The implications of these findings in terms of referral-competition networks between health organizations and between social venues are discussed.

This study examines the co-evolution of conversation and advice networks in a Japanese university class, based on the self-determination theory in which autonomy, competence, and relatedness are the three key concepts. The network data were collected during the spring and fall semesters from 2013 to 2016. Stochastic actor-oriented models were used to estimate the dynamics in the two networks. The results are as follows: (1) There was an asymmetric property in the advice ties; (2) One tie of a network created a tie of another type of network at the dyadic level; and (3) For enhancing advice ties, the autonomy effect of conversation tie was effective in the fall semester. The findings of this study offer a new understanding of the dynamics between conversation and advice networks.

It has repeatedly been examined that occupational evaluations are stable over time and various viewpoints. However, it does not mean that the evaluation will not change at all. This paper examined whether occupational evaluations were changed by the existence of a person who engaged in the particular occupation among the respondents’ family, relatives, friends, or acquaintances. We can think of some mechanisms for the evaluation change. First, we receive some support from a person who engaged in the focal occupation, and give a higher evaluation to the occupation in return. We call this mechanism the “support effect.” Second, we may give a higher evaluation to our own occupation than others do, and this results that we give a higher evaluation to others’ occupation of the same kind. We call this mechanism the “self-enhancement effect.” Third, as we get to know more about the compensation of the focal occupation, we may give a higher evaluation to the occupation. We call the mechanism the “knowledge effect.” Exploratory analyses were conducted to examine which mechanism applies to each occupation. We could find out only a few occupations whose evaluation got higher when we had a person in the focal occupation in our family / relatives / friends / acquaintances; and among those occupations, the support effect was seen in about a half of them. In addition, among those occupations in which knowing someone engaged in them did not make a difference in evaluations, some of them showed support effect.

These days, Japan has seen an explosion of art festivals called “art projects” across the country. The Setouchi Triennale (Setouchi International Art Festival) is a contemporary art festival held every three years on a dozen islands in the Seto Inland Sea. An art festival can be regarded as successful to the extent that it represents authentic regional images well both on tourist and resident sides. We delve into how art experiences help construct visitors’ island images by way of brand associative network analyses. We use brand concept data collected on two islands, Shodoshima (the largest and most populous island in this region) and its neighboring small island, Teshima that has been suffered from an illegal dumping of waste. Art plays a pivotal role in constructing island images for Teshima while it is not the case for Shodoshima that is already filled with island images through long history of sightseeing and film shooting. The art festival also contributes to revealing the illegal dumping of waste in the eyes of tourists through the famous Teshima Art Museum to some extent.

Recent developments in Information Technology have drastically changed and, at the same time, enabled the accumulation of substantial digital data for our social world. An emerging field that explicates the reality and mechanisms of social phenomena using large volumes of digital data is called “computational social science.” This paper overviews the prospects and challenges for computational social science, especially focusing on its application in sociology. First, we define what computational social science means. Second, we explain the reason why this emerging new field has a special relationship with sociology. Third, we clarify where novelty lies in computational social science and discuss its analytical methods. Finally, we consider some challenges in this field.