Cultural Manifold Analysis (CULMAN) as a Paradigm of Cross-national Comparative Surveyson National Character

This paper presents an overview of the history of national character study of the Institute of Statistical Mathematics over the past 60 years. ``Japanese National Character Survey,'' started in 1953, is a rare longitudinal survey in the world of research, and it motivated other countries to start similar longitudinal surveys, such as GSS, ALLBUS, Eurobarometer, etc. Since the early 1970s, the Japanese survey has been extended to a cross-national survey series for more advanced study on Japanese national character in a comparative context. This paper touches on some aspects of those surveys, including, the cross-national studies and the paradigm called ``Cultural Manifold Analysis (CULMAN).'' It is fortunate indeed to encourage a wide range of readers to better understand the ``cross-national comparative surveys of attitudes, opinion and social values'' as basic information for scientific research and policy-making.

Key words: Asia-Pacific Values Survey, cross-national survey, cultural manifold analysis, Japanese national character, statistical random sampling survey, valid response rate.

A Study of Survey Non-response Using the 13th Nationwide Survey of the Japanese National Character: Assessing Sampled Individuals', Neighborhood' and Survey Takers' Characteristics Associated with Non-response Behavior

This study assesses nonresponse behaviors of a large-scale social survey. Using data of the 13th nationwide survey of the Japanese National Character, a survey for survey takers, and national population census of 2010, this study investigates whether sampled individuals', neighborhood', and survey takers' characteristics relate to the individuals' survey response behaviors. As sampled individuals (N=6054) are nested in each sampled neighborhood (N=400), a multinomial two-level multilevel modeling approach was applied. Results of the study indicate that, when compared to a reference category (i.e., response), (1) whether a person lives in a detached home is related to refusal by target persons and by others (i.e., other than sampled individuals, likely family members), (2) there are differences between neighborhoods in survey response behaviors, especially between-neighborhood-variation of non-contact cases is relatively large, (3) neighborhood-level variables obtained from national population census of 2010 partly explain differences between neighborhoods in the behaviors, and (4) survey takers' years of experience are associated negatively with non-contact and refusal by target individuals and by others. The study's findings indicate that more experienced survey takers are capable of avoiding non-contact and refusal cases. Since survey administrators cannot force sampled individuals to respond to a survey, but can select and train survey takers, further studies need to examine relationships between survey takers' characteristics and sampled individuals' non-response behaviors to have practical implications that can improve survey response rates.

Key words: Survey of the Japanese National Character, survey nonresponse, social surveys, survey takers, census, multilevel modeling.

Relationship between Attitudes for Volunteer Participation and Views of Japanese Society: From the 12th and 13th Nationwide Surveys on the Japanese National Character

This study analyzes factors affecting the attitudes of volunteer participation in the 12th and 13th surveys on the Japanese National Character using multinomial logistic (MNL) regression. The MNL regression analyses use data divided into the free competition group and the protection group for the weak, and their results are compared to examine whether and how the factors behind the attitudes of these two groups differ. In both the 12th and the 13th surveys, we find that divided models show a better fit than non-divided ones and that different variables of the models for the two groups are significant. Thus, it is considered that volunteer participation in these two groups is driven by different factors. Moreover, different variables are significant between the 12th and 13th surveys in both groups. In particular, many significant variables differ for the free competition group. This finding suggests that the decision mechanism for attitudes or character of the free competition group has changed over the examined five-year period.

Key words: View of Japanese society, volunteer activity, multinomial logistic regression, Japanese National Character Survey, free competition, welfare state.

Who Feels That Effort Is Rewarded? Analysis about a Feeling of Effort from Survey on the Japanese National Character

Japanese people believe that efforts will be rewarded in the future, and this has been accepted as a virtue for a long time. However, in modern Japanese society, this sense of values over efforts and reward is changing. This study aims at further exploring Japanese attitudes to the relation between efforts and reward, and investigates the factors promoting Japanese feelings that efforts are rewarded. To examine the determinants of ``sense of rewarded efforts'', we conducted logistic regression analysis using data of ``the Survey on the Japanese National Character''. The main findings were as follows: (a) from comparison between the 8th survey in 1988 and the 13th survey in 2013, the view of ``efforts are rewarded'' diminished in the collapse of the bubble economy and long-term economic recession during this period. (b) the view of ``efforts are rewarded'' are affected by not only a perceived inequity in social structural factors, consciousness of social isolation, but also a feeling of attachment to their own country and perceptions of fairness of society.

Key words: The Survey on the Japanese National Character, sense of rewarded effort, social isolation, perceived fairness of society.

An Investigation of Meanings of ``Trust'' and Their Transition Using Latent Class Analysis in Japanese National Character Survey

Trust has long been studied as one of the underlying mechanisms enabling societies to function smoothly. However, there are problems with measuring ``trust'' through social surveys. Because the word ``trust'' has a large variety of meanings, even if we use the same scale for measuring, it may be difficult to capture an identical aspect of it. This suggests that it is not enough to observe trends of response distributions of trust-related items separately. To grasp the actual situations of time trends on trust, this paper investigates the conceptual structure of trust in the Japanese National Character Survey using a latent class model. As a result, we extracted three latent classes, named ``carefulness'', ``generalized trust'', and ``distrust (assurance)'' based on values of conditional response probabilities of the model. Furthermore, we examined the proportions of the three classes in each survey. In 1978, the percentage of ``carefulness'' was the highest of all, but this gradually changed with the times. The percentage of ``carefulness'' has decreased gradually over recent decades. On the other hand, the percentage of ``generalized trust'' increased rapidly in the early 1990s, ultimately reaching around 50% in 2013. We also examined the effect of background variables on class membership.

Key words: The Japanese National Character Survey, trust, latent class model, time transitions.