Cross-National Comparative Survey of National Character

Cross-National Survey Team


Ryozo Yoshino (Professor, Department of Data Science, The Institute of Statistical Mathematics)


Tatsuzo Suzuki (Emeritus Professor, The Institute of Statistical Mathematics)
Fumi Hayashi (Professor, Toyo-Eiwa Women’s University)
Kazue Yamaoka (Chair, Department of Technology, Assesment and Biostatistics, National Institute of Public Health, )
Yuejun Zheng (Professor, Doshi-sha University)
Wataru Matsumoto (Associate Professor, Kansai University)

Cross-National Comparative Survey of National Character

Ryozo Yoshino
Cross-National Survey Team of National Character
The Institute of Statistical Mathematics

1. Introduction

This homepage shows some results on our past surveys including, “A Cross-National-Survey of Seven Countries” (Japan, U.S.A. & 5 European Countries in 1988-1993), “East Asia Value Survey (Japan, Beijing, Shanghai, Taiwan, South Korea, & Singapore in 2002-2005) ” and “Asia & Pacific Value Survey”(2004-2009 fiscal year) by the cross-national survey team of the Institute of Statistical Mathematics (Chief: Ryozo Yoshino).  We are developing this study in order to exemplify practical research of a new methodology for cross-national comparative survey, called CULMAN (Cultural Manifold Analysis) (See Yoshino[ 2005b] and Yoshino, Nikaido and Fujita[2009]).
Some historical background of this study is as follows.  (See Yoshino [2001, 2005c, 2006] and Yoshino & Hayashi [2001] for more detailed English explanation on the history, methodologies such as back-translation technique for constructing questionnaire and statistical random sampling, a paradigm of cross-national comparability.  As for the information on our past surveys, see the series of ISM Research Reports published over decades or our home page of the Institute of Statistical mathematics.  ( )

2. Some History on Our National Character Survey

The Institute of Statistical Mathematics (ISM) has been conducting a longitudinal nationwide social survey on the Japanese national character every five years since 1953, using the same questionnaire items (Mizuno et al., 1992).  The survey is called “Nihonjin no Kokuminsei Chosa” (Japanese National Character Survey). Although definition of the term “national character” may be very problematic, here it simply means the characteristic shown in people’s response patterns to a questionnaire survey (Hayashi et al., 1998; cf. Inkeles, 1997).   The question items cover various aspects of people’s opinions about their culture and daily life.  This survey was one of the foundations of the public opinion survey system based on the statistical sampling theory developed immediately after World War II in Japan.  The significance of this survey was clear at the time when Japan was expected to shift from the military regime to a democratic system in the latter half of 1940s (Yoshino, 1994).  This survey stimulated many countries to carry out the same sort of time series surveys such as the World Value Survey, Eurobarometer, General Social Survey (GSS) of USA, ALLBUS of Germany, CREDOC of France, etc. (There was a time that the post-war Japanese democracy had been criticized because it was not democratic from a viewpoint of the Western world. Interestingly, however, Japan conducts public-opinion polls based on a statistically ideal sampling using an almost complete residential or voters’ list whereas the other countries have to use other methods such as quota sampling or random-route sampling.  The latter two sampling methods consider statistical randomness but do not yield the statistical estimate of sampling errors.  As far as the system of public-opinion polls is concerned, therefore, Japan may be more democratic than the Western countries.)
Since 1971, the survey of ISM has been extended to a cross-national comparative study for more advanced understanding of Japanese national character (Hayashi, 1973). The focus of our cross-national surveys is the investigation of the statistical comparison of peoples’ social values and their ways of thinking and feeling.  More explicitly, our concern has been with cultural identities and people’s attitudes toward economy, freedom of speech, interpersonal relationships, leadership, politics, public acceptance of science and technology, religion, social security, etc.  These aspects may clarify certain similarities or dissimilarities that are represented by psychological distances between countries or races in certain statistical pattern analyses of responses (Hayashi, 2001a, 2001b; Hayashi et al., 1998; Yoshino, 1994, 2001c).

Table 1 .List of Our Past Surveys on National Character.
1953 Present (every five years).  Japanese National Character Survey
1971 Japanese Americans in Hawaii
1978 Honolulu Residents (including Japanese American) & Americans in the USA (mainland)
1983 Honolulu Residents (including Japanese American)
1987 Britain, Germany & France
1988 Honolulu Residents (including Japanese American)
  Americans in the mainland of U.S.A, the Japanese in Japan
1991 Japanese Brazilians (JB) in Brazil
1992 Italy
1993 The Netherlands
1998 Americans with Japanese ancestry in the West coast of U.S.A.
1999 Honolulu Residents in Hawaii
2001 China 2001 Survey (Beijing, Shanghai, Kunming, Hangzhou)
East Asia Value Survey(2002-2005 fiscal year)
2002 Japan 2002 survey, China 2002 Survey (Beijing, Shanghai & Hong Kong)
2003 Taiwan 2003 Survey, South Korea 2003 Survey
2004 Singapore 2004 Survey
Pacific-Rim Value Survey (2004-2009 fiscal year) *
  Japan 2004A survey & Japan 2004B survey
2005 China 2005 Survey (Beijing, Shanghais & Hong Kong)
2006 Taiwan 2006 Survey, South Korea 2006 Survey, USA 2006 Survey
  (All the above-mentioned surveys were based on statistical random sampling data.)
2007 Singapore 2007 Survey, Australia 2007 Survey
2008 India 2008 Survey

* Note that this survey series was formerly called the Asia-Pacific Value Survey. However, because we have elected to designate the newest and currently running set of survey projects (fiscal years 2010-12) with that title, accordingly the 2004-09 surveys have been renamed as Pacific-Rim Value Survey.

The cross-national survey, however, involves particular methodological problems.  It is not simple to compare response data collected under different conditions. Different countries may use the same questionnaire but in different languages and employ different statistical sampling methods as well.  There is no a priori knowledge as to how these different conditions influence peoples’ responses even in the case where there is no substantive difference of opinions and social values between peoples (Yoshino, 2001c).  Thus, an important problem of our study is to investigate those conditions under which meaningful cross-national comparability of social survey data is guaranteed. As our approach towards this problem over decades, we have been developing the methodology called CLA (cultural link analysis).  The main components of CLA are 1) a spatial link for cross-national comparison, 2) a temporal link inherent in longitudinal analysis, and 3) an item-structure link inherent in the commonalties and differences in item response patterns within and across different cultures (cf. Guttman, 1972).  In CLA we utilize, for example, the back-translation technique and statistical pattern analyses such as Hayashi’s Quantification Method (Hayashi, 1992) or Yoshino’s (1992a, 1992b, 1994, 2001c) Super-culture Model. The utilization of those pattern analyses consists of an important part of our methodology. Namely, although a simple cross-national tabulation of people’s responses with respect to a single item may not be reliable because people’s responses may occasionally be sensitive to slight differences in the wording of certain questions, certain pattern analyses or scaling on a set of items can be reliable. (See Yoshino & Hayashi [2002] for overviews on our approach.)
On the other hand, in this cross-national study, we have found some response tendencies particular to certain countries. For example, the Japanese tend to avoid polar answer categories and to choose intermediate categories, whereas the French generally tend to give negative responses to any question. (Here I may be exaggerating these tendencies to make the points clearer.)  I think that we should consider these response tendencies when we analyze not only people’s sense of trust but public opinion polls or social survey data in general.
See Hayashi (2001a, 2001b), Hayashi et al. (1998), Yoshino (1994, 2001c, 2002, 2005, 2006) and Yoshino & Hayashi (2002) for results of our cross-national surveys.

3. Japanese national character survey (1953-present)

Our longitudinal survey of Japanese national character shows some stable aspects of attitudes and social values of the Japanese (Hayashi & Kuroda, 1997; Yoshino, 1994). Among others, the stability of interpersonal attitudes and religious attitudes may distinguish the Japanese from other countries.  Namely, the Japanese show a higher score on the “Giri-Ninjyo scale” than the other countries.  Moreover, while only one third of the Japanese have religious faith, but more than 60% of the Japanese support the opinion that religious attitudes are important (Yoshino & Hayashi, 2002; Yamaoka, 2000).
I will briefly explain certain fundamental dimensions of the Japanese social values as follows.
Fundamental dimensions of the Japanese social values
Hayashi (1993) has identified two important dimensions that underlie the Japanese national character in the survey. That is, 1) the dimension of interpersonal relationships (“Giri-Ninjyo” attitude, or a complicated sense of humanity and obligation that is particular to the Japanese interpersonal relationships) and 2) the dimension of a modern-traditional contrast in their way of thinking.  On one hand, as mentioned before, the Japanese interpersonal attitude has been stable, at least over the last half century, and probably for much longer than our longitudinal survey.  This corresponds to the first dimension.  On the other hand, for over 100 years since the Meiji Restoration in 1868, Japan has been doing her best to overtake Western science and technology and to develop it into a Japanese adaptation.  Probably this enduring effort has underlined the dimension of the Japanese traditional vis-à-vis a modern orientation in the Japanese way of thinking.
However, the Japanese way of thinking has been gradually changing, and there appeared a generation gap between people of 20-24 years old and those older than 25 years in our survey of 1978 (note that the younger generation as born more than 10 years after the end of World War II. In 1956, the economic white paper declared, “Japan is no longer in the post-war condition,” and this symbolized the start of the high-speed development of industry and economy.  On the other hand, however, Japan had to face many social problems concerning pollution because of the high-speed industrialization around 1970. Since the signs of the younger generation’s changes appeared as early as 1978, their current way of thinking has become more complicated than ever.
Furthermore, the Japanese have been in the confusion of the transition period from the established social system to a system of a highly advanced information age. In this confusion, a Central Research Services, Inc. (2000) survey reports of the majority of Japanese people’s distrust toward traditional systems such as banking, bureaucracy, as well as of congressmen, police, etc., in spite of the stereotype of the Japanese as a highly trustful nation (Fukuyama, 1995).

 4.  The World as a Cultural Manifold

The 20th century was the time of expansion of western civilization.  Differences of cultures occasionally prevented us from our understanding each other.  In this time of globalization, I would like to emphasize the fact that there are various ways of successful social development; therefore, we should not impose one’s own social value on any other country if we intend to develop a peaceful world.
The globalization necessarily changes some institutional systems and customs towards more universal ones under the influences of transnational exchange or trade.  On the other hand, some other systems are becoming more and more sensitive to cultural differences, as a reaction to the globalization. 
In order to facilitate the mutual understanding between the East and the West, we need to keep in mind the differences of social values between them.  The study on the scale of trust (Yoshino, 2005, 2006, 2009) may caution us on the applicability of a certain “single” scale invented by the Western cultures to the Eastern cultures, or vice versa.  For example, it is not always the case in Asia that “the distrust is a culture of poverty” as Banfield (1958) once mentioned.  A Chinese proverb says that “Fine manners need a full stomach” (or “The belly has no ears”), but another says “Be contended with honest poverty.”  Gallup (1977, p.461) reported that they could not find a very poor but still happy people in their global survey.  I think that they missed the reality.  For example, Brazilians were very optimistic even when Brazil fell down to the worst debtor nation in the world (Inkeles, 1997). Inglehart reported a correlation of .57 between economic development and life satisfaction for some 20 countries surveyed in 1980s (Inkeles, 1997, p. 366-371).  But the life satisfaction of Japan in the 1980’s was lower than around 2000, although Japan was close to the top of the world economy in those days and now she has been suffering from depression over years until recently. Thus, we need scales constructed from various perspectives or social values in order to understand various cultures in the age of globalization.
Although China had so many battles between small countries (within the area corresponding to the modern China) over thousands of years in their history, once they were synthesized as a large empire, their government employed peoples of various races as high class bureaucrats.  This made it possible for them to develop and maintain a large empire and their culture, often over centuries.  This is analogous to the Roman Empire, but it is contrastive to the modern western countries (and Japan during WWII) that colonized Asian and African countries in the 19th and 20th centuries.  The history shows that trust between different races changes according to social conditions in the long run, although it is relatively stable over time.
After our previous China survey (China 2001 survey [Yoshino, 2006]), there occurred the problem of SARS spreading from Guang-Zhou in China. People inside and outside China criticized the local governments, suspecting that they attempted to hide the serious conditions. This seems to suggest a significant change of China, from secretive attitude to more open attitude for every matter. The secretive attitude was linked to the system of severe punishment on political responsibility. The open attitude is a key to democracy that is necessary for successful capitalism. The mayor of Beijing got fired because of his mishandling of SARS. The government started encouraging people to inform of the presence of patients. This situation seems to show that China is changing rapidly, but in a Chinese way. Here it may be important to quote Dogan (2000)’s statement“... Erosion of confidence is first of all a sign of political maturity. It is not so much that democracy has deteriorated, but rather the critical spirit of most citizens has improved ...” In spite of prevailing confusion in East Asia (actually in the entire world), I hope that East Asia will advance towards the peaceful development without serious conflicts. For the mutual understanding among Asian countries, one should keep in mind their ways of thinking such as “Mentsu (face)” and “Honne and Tatemae (real intension and principle)” of the Chinese, the Japanese, and the Korean. This is also the case with the Asian countries for their understanding of the West. Once upon a time, Weber (1904-05) argued that Asian countries were not able to develop capitalism in his theory on religion and capitalism. Now we know so many counter-examples such as Japan, Korea, NIES, and China, against his argument. Some people argued that the Japanese adaptation of Confucius philosophy adapted to Japan functioned as a replacement of Protestant ethics and led Japan to a successful development of capitalism (Morishima, 1984). But the past decades have seen many examples to show that economic success is not linked to a particular religion. Now we have more and more data to consider the relationships between economic development, social systems and social values because of the rapid change of social systems in many countries of the world than before. I hope that our survey data will be helpful for further constructive arguments, and the mutual understanding for the peaceful development of Asia.


Dogan, M. (2000).  Deficit of confidence within European democracies.  In M. Haller (ed.), The making of the European union (pp.243-261).  Springer-Verlag: Paris.
Fujita, T., & Yoshino, R. (2009). Social Values on International Relationships in the Asia-Pacific Region. Behaviormetrika, 36, 2, 149-166.
Hayashi, C. (Ed.) (1973).  A study of Japanese-Americans in Honolulu, Hawaii (in Japanese).  ISM Research Report General Series, No.33.  ISM: Tokyo.
Hayashi, C. (1992).  Suuryou-ka [Hayashi’s quantification method].  Asakura-syoten.
Hayashi, C. (2001).  De-ta no kagaku [The Science of Data].  Asakura-syoten: Tokyo.
Hayashi, C., Yoshino, R., Suzuki, T., Hayashi, F., Kamano, S., Miyake, I., Murakami, M., & Sasaki, M. (1998).  Kokuminsei nanaka-koku hikaku [cross-national comparison of seven nations].  Idemitsu-syoten: Tokyo.
Inkeles, A. (1997).  National character.  Transaction: New Brunswick.
Mizuno, K., Suzuki, T., Sakamoto, Y., Murakami, M., Nakamura, T., Yoshino, R. (1992).  Dai 5 nihonjin no kokuminsei (Japanese national character V) (in Japanese).  Idemitsu syoten: Tokyo.
Sakamoto, Y., Nakamura, T., Tsuchiya, T., Maeda, T., & Fouse, D.B. (Eds.) (2000).  A study of the Japanese national character: the tenth nationwide survey (1998).  ISM Research Report, General Series, No. 85.
Yoshino, R. (1992a).  Superculture as a frame of reference for cross-national comparison of national characters.  Behaviormetrika, 19, 1, 23-41.
Yoshino, R. (1992b).  The unbiased BIGHT model and its application to the analysis of responses of national characters.  Behaviormetrika, 20, 2, 171-186.
Yoshino, R. (Ed.) (1995).  A handbook of cross-national comparative survey.  ISM Research Report No.77.  ISM: Tokyo.
Yoshino, R. (Ed.) (2001a).  A study of statistical science on cultural transmission based  on social survey data of National Character.  A report submitted to The Ministry of Education, Culture, Sports, Science & Technology.
Yoshino, R. (Ed.) (2001b).  Hawaii Resident Survey 1999-2000.  The ISM Research Report, No.86.  ISM: Tokyo.
Yoshino, R. (2001c).  Kokoro wo hakaru [Measuring the mind].  Asakura-syoten: Tokyo.
Yoshino, R. (2002a).  A time to trust --- a study on peoples’ sense of trust from viewpoint of cross-national and longitudinal survey on national character ---. Behaviormetrika, Vol.29, No.2, 231-260.
Yoshino, R. (2003).  Shinrai no toki (A time to trust) [in Japanese].  Eco-Forum, Vol. 22, No.1, a special issue on social capital II, pp.42-51.
Yoshino, R. (2005a).  Higashi-Ajia kati-kan tyousa –bunnka tayoutai-kaiseki ni motozuku keiryo-teki bunmeiron koutiku ni mukete- (East Asia Value Survey – For the development of Behaviormetric study of civilization on the Cultural Manifold Analysis (CULMAN) [in Japanese].  Japanese Journal of Behaviormetrika, Vol. 32, No.2, pp.133-146.
Yoshino, R. (2005b).  A time to trust in the East Asia – A Behaviormetric study on the sense of trust in East Asia Value Survey- [in Japanese].  Japanese Journal of Behaviormetrika, Vol. 32, No.2, pp.147-160.
Yoshino, R. (2005c). Trust and National Character--- Japanese sense of trust,  Cross-national and longitudinal surveys-.  Comparative Sociology, Vol.4, No.3-4, pp.417-450.
Yoshino, R. (2006).  A social value survey of China --- on the change and stability in the Chinese globalization ---.  Behaviormetrika,Vol.33, No.3, pp.11-130.
Yoshino, R. (2009). Reconstruction of trust on a Cultural Manifold: Sense of Trust in Longitudinal and Cross-National Surveys of National Character. Behaviormetrika, 36, 2, 115-147.
Yoshino, R. & Hayashi, C. (2002b).   An overview of cultural link analysis of national character.   Behaviormetrika, Vol.29, No.2, 125-141.
Yoshino, R., Nikaido, K., & Fujita, T. (2009). Cultural Manifold Analysis (CULMAN) of National Character: Paradigm of Cross-National Survey. Behaviormetrika, 36, 2, 89-113.
Zheng, Y. (ed.) (2002).  Researches on the national character of Chinese and Japanese--- A sampling srvey in Beijing, China ---.  ISM Research Report No.89.
Zheng, Y (2003).  Researches on the national character of Chinese and Japanese--- A sampling survey in Shanghai, China ---.  ISM Research Report No.90.
Zheng, Y. & Yoshino, R. (2002).  The sampling methods for the China survey (Beijing and Shanghai) [in Japanese].  Proceedings of the 30-th conference of Japan Behaviormetric Society, pp.346-349.
Zheng, Y. & Yoshino, R. (2003).  For the comparative study of East Asia value survey (in Japanese).  The Japanese Journal of JAPOR [Japanese Association of Public Opinion Research], Vol.91, pp.16-21.
Tsunoda, H., Yoshino, R., & Yokoyama. (2008). Components of Social Capital and Socio-Psychological Factors That Worsen the Perceived Health of Japanese Males and Females. The Tohoku Journal of Experimental Medicine, Vol.216, No.2,pp.173-185.
Yoshino, R., Nikaido, K., & Fujita, T. Cultural manifold analysis (CULMAN) of national character: paradigm of cross-national survey. Behaviormetrika, Vol.36, No.2, pp.89-114. (2009)
Fujita, T., & Yoshino, R. Social values on international relationships in the Asia-Pacific region. Behaviormetrika, Vol.36, No.2, pp.148-165. (2009)
Behaviormetrika, Vol.36, No.2, pp.148-165.
Hayashi, F. & Nikkaido K. (2009). Religious Faith and Religious Feelings in Japan: Analyses of Cross-Cultural and Longitudinal Surverys. Behaviormetrika, Vol.36, No.2, pp.167-180.

   For the other English reference, see the list of reference in Yoshino (2002, 2005c, 2006)& Yoshino & Hayashi (2001), which are available as a free online journal of Behaviormetrika, except Yoshino (2005c):

As for the Japanese references, see the list of reference in Yoshino (2005a, 2005b, 2006)& Yoshino & Hayashi (2001), which are available as a free online journal, Japanese Journal of Behaviormetrika:

吉野諒三.  (2003). 「信の崩壊」---世論調査方法論の今日の課題. 行動計量学. 展望「21世紀の行動計量学」 第29巻第1号, pp.45-54.
吉野諒三. (2003)「信頼の時代」. Eco-Forum, Vol.22, No.1, 特集号「ソーシャル・キャピタルPart II」,pp.42-51. 統計研究会.
吉野諒三. (2005). 東アジア価値観国際比較調査―文化多様体解析(CULMAN)に基づく計量的文明論構築へ向けて―. 行動計量学. 第32巻2号, pp. 133-146.
吉野諒三. (2005). 富国信頼の時代へ―東アジア価値観国際比較調査における「信頼感」の統計科学的解析―. 行動計量学. 第32巻2号, pp. 147-160.
吉野諒三、鄭躍軍、朴承根.  (2003).「東アジア諸国の人々の日本語観.」 行動計量学, 第30巻 第1号(通巻58号), pp.311-52.
吉野諒三. (2005).東アジア価値観調査-文化多様体解析(CULMAN)に基づく計量文明論の構築へ向けて —.  行動計量学. Vol.32, No.1, pp.133-146.
吉野諒三.(2005).富国信頼の時代へ—— 東アジア価値観国際比較調査における信頼感の統計科学的解析 ——.  行動計量学. Vol.32, No.1, pp.147-160.
三好 美浩, 吉野諒三. (2005). 東アジアの職業観—— 日本・中国・台湾・韓国の比較 ——. 行動計量学. Vol.32, No.1, pp.173-189.
鄭躍軍, 吉野諒三、村上征勝. (2006). 環境意識 —— 日本・中国の比較 ——. 行動計量学. Vol.32, No.2, pp.55-68.
袰岩晶・吉野諒三・鄭躍軍 (2008). 国際比較における「データの安定性」に関する一考察---中国調査データの検討を通した文化多様体解析の試行---. 統計数理、55,2,pp.285-310. 
(行動計量学32巻2号、33巻1号及びBehaviormetirka、Vol.29, No.2、Vol.30, No.1、Vol.36, No.2、Vol.37, No.1の特集号も参照していただきたい。)