Modernization Theory Revisited: A Cross-Cultural Study of Adolescent Conformity to Significant Others in Mainland China, Taiwan, and the USA
Jie, Zhang, Thomas, Darwin L., Adolescence
Modernization theory can be traced to the writings of Emile Durkheim (1987) and Max Weber (1992). Modernization is a broad concept that refers to major social changes which occur when a preindustrial society develops economically and the workplace shifts from the home to the factory (industrialization), people move from farms into cities where jobs are available (urbanization), and large-scale formal organizations emerge (bureaucratization). These three components of modernization accompany changes in such major asocial institutions as the family, religion, and education. These changes in turn affect power relations among people in a society.
Modernization theorists in social psychology have established a relationship between economic advancement and the power structure. Many previous studies, including those of Inkeles (1972) and Thomas and his associates (1971, 1972, 1974), have found a decrease in traditional face-to-face social relations and an increase in secondary social relations as industrialization and urbanization increase. These studies established that as the modernization level increases, traditional authoritative others such as leaders in the family, religion, and community have less power because they control fewer resources, while those in education, employment, and media control more resources and have greater power.
However, much of the previous modernity research has treated culture in a problematic fashion, failing to measure it adequately or to systematically analyze for cultural differences. Thus the effect of culture in a society on power and conformity has not yet been fully specified. Although Thomas and his associates dealt with multiple cultures, culture in their studies was confounded with modernization level. As a result, cultural effects could not be easily identified. Since culture to a society is as personality to an individual, cultural differences are very important in studies of power and conformity. For instance, if a society values education highly such as in the tradition of Chinese culture, modernization will hardly result in much increase in the youths' conformity to educators. Therefore, the study of the effect of modernization on power relations needs to include the cultural variable.
The dependent variable of this study is adolescent conformity to significant others. From modernization theory we begin with the assumption that the higher a society ranks on the industrialization-urbanization continuum, the lower the tendency of its youths to conform to the expectations of traditional significant others (such as parents and religion leaders), and the higher the tendency to conform to rational/legal significant others (such as professors). It is further hypothesized that culture will modify adolescent conformity patterns predicted by modernization theory.
The first purpose of this research was to test for hypothesized differences in adolescent conformity behavior in three different societies varying in their level of modernization and culture: the United States, Taiwan, and mainland China. Since these three societies are on the urbanization-industrialization continuum from high to low,(1) and the culture of the U.S. is different from that of the two oriental societies,(2) a comparison of conformity patterns among different levels of modernization across cultures is possible. Hypotheses of this research are illustrated in Table 1.
Taking the first significant others, parents, as an example, modernization theory predicts that the more modernized the society, the less power parents have, since they control fewer resources in such a society. But when we consider the cultural tradition of Confucianism in Taiwan and mainland China, we expect no significant difference between these Chinese societies on adolescent conformity to parents. Modernization theory posits that religion will lose its influence in society as a result of secularization, but the cultural values that emerged in mainland China after the communist takeover will result in less influence for religion. Modernization theory also predicts greater influence of the educator. Because of the cultural background of Confucianism, the Far Eastern societies will not score lower in conforming to the educator than will American society, even though the Chinese societies, especially mainland China, are less modernized (Smith, 1991). In general, this research allows for a more careful assessment of how culture modifies and illuminates modernization theory.
The second purpose of this research was to develop a LISREL path model to test the influence of several variables on youths' conformity to significant others. It has been documented that gender (Milgram, 1974; McGuire, 1985), family socioeconomic status (Lee, 1982; Kohn, 1976), family integration (Rollins & Thomas 1979), and religiosity (Weigert & Thomas 1979); Chusmir & Koberg 1988), play important TABULAR DATA OMITTED roles in youths' conformity to significant others. It is presumed that people of different cultures have in common, to a large extent, their conformity to authorities (Milgram, 1961, 1963, 1964). By studying the above-mentioned independent variables together, a pattern of youths' conformity to authorities can be established in terms of which variables are more important in predicting level of conformity within and across cultures.
The literature has indicated a strong relationship between gender and conformity, with females more likely to conform to traditional significant others (Milgram, 1974; Eagly, 1978; Tuthill & Forsyth, 1982; McGuire, 1985). Although some studies failed to show significant gender differences (Maccoby & Jacklin, 1974; Cooper, 1979; Eagly & Chrvala, 1986), the majority indicated that women tend to be more conforming and more susceptible to social influences. In the relation between gender and religiosity, we know that females are usually more religious than males, and this is expected to hold true in all three cultures.
The relationship between family socioeconomic status and conformity is not simple. Higher social conformity was found among youths from the middle SES group than among those from low and high SES backgrounds (Singh, 1984). This psychological issue is related not only to family economic status, but to parents' occupations and their expectations for their children, which are determined by family economic status (Kohn, 1959a, 1959b, 1963, 1976). However, other studies have reported a negative correlation between family SES and youth conformity to traditional significant others in the form of compliance (Barry, Child, & Bacon, 1959). In addition, in the model, family SES is expected to be positively correlated with adolescents' academic behavior, since the literature has indicated that the higher the family social economic status, the higher the students' academic achievements (Coleman, 1966; Jencks, 1972).
Family integration, which represents both family cohesiveness and parental support, not only increases adolescents' traditional conformity, but also influences their academic behavior and their religiosity (Rollins & Thomas 1979). It is hypothesized here that the higher the family integration, the higher the level of conformity of adolescents to traditional significant others.
There are confusing results in the studies of the relations between religiosity and conformity. Some studies show that the two variables are positively correlated (Albrecht, Chadwick, & Alcorn 1977), while others indicate that there is no correlation (Stark, 1984). However, most other studies indicate that religiosity varies directly with conformity to traditional significant others (Weigert & Thomas, 1979; Rigby & Densley, 1985). In this path model, parents' religiosity is hypothesized to be strongly related to one's own religiosity and also to academic aspirations and performance. Although no literature is available regarding the relation between religiosity and media/peer conformity, a negative correlation is hypothesized here because the authors believe that those who are more integrated with their family members will be less conforming to their peers.
Since the resource theory of power (Collins, 1974) tells us that those students who do well at school and highly value grades and knowledge will allocate more power to educators who control valued academic resources, it is hypothesized that students with higher academic aspirations and performance are more likely to conform to the expectations of the educator.
Gender, family, SES, family integration, parents' religiosity, academic behavior, and self religiosity are obviously predictors of adolescents' conformity to significant others. But it is not yet known how these variables rank in terms of their predictability. Additionally, some factors are likely to influence adolescent conformity both directly and indirectly (such as gender, family SES, and family integration), while others such as parents' religiosity may influence conformity through different means. In order to compare them, a comprehensive path model is necessary. Figure 1 illustrates the hypothetical model.
Weber's typology of authorities was followed in structuring the dependent variable of the model. Significant others are of three types: traditional, rational/legal, and media/peer (Weber's "charismatic" is replaced by media/peer, since in today's world charisma can be applied to almost any figure). Traditional significant others are defined as father, mother, and religion leader. Rational/legal is professor, and media/peer are represented by entertainment star and best friend. The model includes three dependent variables in order to permit estimates on each of the three types of significant others. Since this research focuses on the variations between the three cultures, the model is treated within each culture.
The variables of gender, family SES, family integration, and parents' religiosity are set as social environment variables. Adolescents' academic behavior (a combination of expectation and performance), and adolescent religiosity are subjective dimensions of the youths' world and are conceptualized as intervening between the social structural variables and adolescent conformity.
The nature of this model is informed by structuralist sociologists such as Marx and Durkheim. The model indicates that conformity is generally determined by both social environments and individual traits, and individual characteristics are influenced by social factors. Therefore, social factors are frequently supposed to influence adolescent conformity through personal factors.
The hypothetical model proposes relations between the variables. On each path line, both the direction and nature of the relation (positive or negative) are provided in accordance with theories and previous studies. In the dummy variable of gender, "male" is set to "zero" and "female" is set to "one." Since it is hypothesized that females are more conforming than males, the path line between gender and conformity is also positive.
Population and Sampling
The target population of this research is late adolescents (around 20 years old) from mainland China, Taiwan, and the U.S., varying both in culture and in modernization level.
Only college freshmen (with a small number of sophomores) were surveyed. The reason freshmen in higher learning institutes were used is that Chinese and Taiwanese freshmen should be less westernized or modernized than their senior counterparts, and American first-year students are less sophisticated about different cultures than are upper-level students. The decision to use college students rather than high school students or nonstudent adolescents was based on two considerations: (1) college students are informed about the educational, occupational, political, family, and religion realms and can respond to the questions; and (2) the limited cost of sampling college students made the research possible. The sample was not random due to geographical restraints, the political situation in China, as well as the budget.
Five universities for the survey were selected through informal social networks: three in the U.S., one in Taiwan, and one in mainland China. The U.S. universities are located in three different midwestern and western states. The university in Taiwan is in Taipei, and the university in mainland China is in Beijing. The schools vary greatly in level of modernization. Respondents from the U.S. universities have been living in a modernized society, and although the schools in mainland China and Taiwan are in metropolitan areas, they are not on the same level of modernization as the U.S. schools. Respondents from Taiwan have been raised in a comparatively more modernized society than those in Beijing.
Convenience samples were drawn from classes in which the instructors granted time for the survey. The total number of valid cases for analysis was 1,026--296 from mainland China, 362 from Taiwan, and 368 from the U.S. The mean age was 19 years, and the total sample is slightly overrepresented by females (55.8%).
The techniques used in constructing the dependent variables were derived from previous research (Thomas, Gecas, Weigert, & Rooney, 1974). This instrument used places the respondent in a hypothetical situation in which he or she is asked to resolve a dilemma by doing what "you want to do" in each of the specific situations or else by doing what the significant other "wants you to do," (Thomas et al., 1974). The following illustrates the situational dilemmas:
Suppose that your father thinks that a particular coat looks good on you and that you should buy it. You do not like it. What would you do? Refuse to buy the coat: (a) absolutely certain; (b) fairly certain; (c) I guess so. Or: buy the coat anyway: (d) I guess so; (e) fairly certain; (f) absolutely certain.
Each of the questions asks the respondent to make a decision as to whether to conform to or refuse the expectation of a significant other. The decision must be made on a six-category scale, from total refusal to complete conformity. The conformity items are used as dependent variables in both ANOVA tests and LISREL path models.
The concept of family socioeconomic status includes four items: (1) the total income before taxes of the family in the last year. (There are seven categories from less than $1,000 to $25,000 or over; the amounts for mainland China and Taiwan respondents are modified in accordance with their income situations and currencies.); (2) subjective evaluation of the family's economic status compared with others in that locale. (There are four categories: the bottom 20%, below average, above average, and the top 20%); (3) and (4) parents' educational level. (There are 10 categories in each, from less than elementary school to completed graduate school.)
The latent variable of family integration has five indicators: (1) the extent to which the respondent enjoys family activities. (Five categories ranging from not at all to all the time); (2) and (3) how well each parent knows the respondent's best friends. (Five categories ranging from not at all to all of them very well); (4) and (5) the extent to which the respondent talks with each parent about problems. (Five categories range from never to always).
Parents' religiosity is measured by six items (1) and (2) the respondent's subjective judgment of each parent's religiosity. (There are four categories for each, ranging from not at all religious to very religious.); (3) and (4) number of times in one week each parent prays. (There are 10 categories in each ranging from none to more than seven times a week).; (5) and (6) each parent's church attendance. (There are eight categories in each from never to more than once a week.)
Academic aspirations and performance is measured by three items: (1) grade expectation for college education (Eight categories range from D's and F's to mostly A's.); (2) grades received so far. (There are also eight categories ranging from D's and F's to mostly A's.); (3) the importance of good grades to the respondent. (Five categories range from not important at all to extremely important.)
Measures of religiosity of the respondent include four items which are basically the same scales with one addition as used in parents' religiosity: (1) self judgment of religiosity; (2) closeness to the Deity. (Five categories ranging from do not believe in God to extremely close to God.); (3) number of times a week to pray; (4) church attendance.
Translation and Back Translation of the Questionnaire
In order to avoid or reduce nonequivalent measurements in the English and Chinese versions, after the questionnaire was translated into Chinese, the back translation technique was employed (Thomas & Weigert 1972); Thomas et al., 1974; Lee, 1984).
The first author, a native speaker of Chinese, first translated the questionnaire into Chinese. The English version was then translated again into Chinese by three bilinguals who speak Chinese as their first language and English the second. The three translations were then compared with one another as well as with the author's for discrepancies. Whenever any differences were discovered, the three bilinguals and the author discussed them and made changes in the translation until there was agreement.
The improved Chinese version was then back translated into English using a similar procedure. Five bilinguals with English as their first language were asked to translate the Chinese version into English. They were selected from American students who were fluent in Chinese. They were unaware of the nature of the study or the original English version. The five English versions were then compared with the original English version. All discrepancies were discussed, and problems resolved.
Although people in Taiwan and mainland China share the same language--Mandarin Chinese, the 40-year separation has split standard Chinese into two different dialects. Besides the obvious differences in the character writing, there are a large number of words which are used and understood differently by the people living on the two sides of the Taiwan Strait. The Chinese version, after being converted into the traditional character form, was read by three Taiwanese college students who had just come to the U.S. They were asked to go through the questionnaire word by word to find any expressions which would prevent Taiwanese respondents from correctly understanding them. Corrections were made after the agreement was reached.
Analyses of variance were employed to determine differences in adolescent conformity between the cultures with different levels of modernization. Each significant other with its items was taken as a unit, and the difference among the three cultures in their conformity to the significant other were studied to test the three hypotheses. Table 2 illustrates conformity means of the three cultures to each of the three significant others as well as the ANOVA results.
The means for each significant other are actually means of the item means. Table 2 provides the variance information of the conformity variables. With the large standard deviation for each mean, there are adequate variances in the conformity variables in each sample to make a comparison across samples possible.
Although analyses of variance indicate significant differences among each of the six significant others, the means are so close, we could not determine if there is a real difference between each pair. ANOVA tests of the SPSSx program were utilized to determine the difference between each pair.
Table 2 Means of Conformity to Each Significant Other: A Comparison of the Three Cultures Modernization Level LOW MIDDLE HIGH SIGNIFICANT M. China Taiwan USA OTHER ETA F Val. Parents MEAN 2.0709 2.1360 1.7332 .25 34.27 S.D. .7039 .6279 .7598 Religion MEAN 1.9501 2.0638 2.3616 .16 13.59 S.D. 1.0661 .9614 1.1485 Educator MEAN 2.8361 2.9229 2.7275 .11 5.74 S.D. .7990 .7090 .8291 Total N = 296 362 366 1024 RELATIONS BETWEEN EACH PAIR Parents Relig. Educa. C and T NS NS NS C and A * * NS T and A * * * * = significantly different at the .05 level NS = not significantly different at the .05 level C = Mainland China T = Taiwan A = the USA
Analysis of variance depicted in Table 2 for conformity to parents reveals an F value of 34.27 and a significance level of .000. A significant difference is shown among the three cultures in youths' conformity to father. However, although the mainland Chinese youths (mean = 2.0709) seem to conform less than the Taiwanese youths (mean = 2.1360), the ANOVA test indicates that there is no statistically significant difference between the two areas in their youths' conformity to parents. The same ANOVA test shows significant differences between Taiwan and the U.S. as well as between mainland China and the U.S. This confirms Hypothesis 1 that "Conforming to parents' expectations, American youths will score lower on the conformity scale than their counterparts in Taiwan and mainland China, but the Taiwanese youths will not score lower than the mainland Chinese youths."
Differences in conforming to religion leader are also presented in Table 2. As hypothesized, the mainland Chinese are the least religious or conforming among the three, although the mainland Chinese (mean = 1.9501) differ significantly only from Americans (mean = 2.3616). Also as predicted from Hypothesis 2, the Taiwanese (mean = 2.0638) conform significantly less than do Americans to religion leaders.
In conforming to educator's expectations, the current data indicate that the Taiwanese (mean = 2.9229) score the highest and the Americans (mean = 2.7275) score the lowest. This point confirms Hypothesis 3 which says that "Students in the two Chinese societies will conform more to their teachers' expectations than will American students . . ." But the second part of Hypothesis 4 is not supported by the data: The Taiwanese students do not conform significantly more to their teachers than do the mainland Chinese (mean = 2.8361).
In summary, ANOVA tests completely confirmed Hypothesis 1, and basically supported Hypotheses 2 and 3.
As indicated, the theoretical model consists of four exogenous and five endogenous latent variables. Independent factor analyses were run to refine each variable by screening and selecting items. This prior examination of the data is known as a specification search and is considered to be a justifiable preliminary step to a confirmatory structural equation analysis (Long, 1983). Only the items that theoretically and/or statistically best represent the latent variable are chosen. LISREL operation further refined the measurement model.
Preliminary LISREL operation indicated that the general model does not fit well in each of the three samples. Family SES and the two religiosity variables were removed from mainland China and Taiwan models, and Academic Behavior was removed from the U.S. model because these variables do not contribute to our understanding of adolescent conformity in the models. Table 3 illustrates the functional variables, their indicators, as well as squared multiple correlations for each of the three samples.
The squared multiple correlations in this table tell us that most of the item variances are well accounted for by their latent variables.
In creating the best fitting LISREL measurement models for the three subsamples, the same procedures were followed as in running the general model. After selecting the final set of beta and gamma parameters with the help of t values, certain lambda X and lambda Y parameters are freed according to the explanation principle. Then, patterned theta delta and theta epsilon parameters are freed until the best possible models are achieved.
Results of the structural models for the mainland China, Taiwan, and the U.S. samples are presented in Figure 2.
LISREL provides several major approaches to test the model. The goodness of fit index is a measure of the relative amount of variance and covariance accounted for by the model and is independent of the sample size. In other words, the overall goodness of fit between the model and the data indicates the closeness of the model to the data. The higher the value of goodness of fit, the closer the model to the data. In these models, the goodness of fit index are .942 for mainland China, .924 for Taiwan, and .894 for the U.S.
Table 3 Squared Multiple Correlations between the Latent Variables and Their Indicators for the Three Samples Exogenous Latent r squared Variables Indicators M.China Taiwan USA GENDER v1 gender 1.000(**) 1.000(**) 1.000(**) FAMILY SES v13 income .130(**) v39 ranking .139 v7 mom educ .434 v8 dad educ .635 FAMILY COHESION v17 enjoy act. .311(**) .086(**) .282(**) v18 mom know fr. .466 .799 .334 v20 talk to mom .438 .317 .525 v21 dad know fr. .598 .503 .841 v23 talk to dad .634 .779 .234 PARENTS' REL. v30 mum relig. .575(**) v31 dad relig. .597 v34 mom pray .759 v35 dad pray .770 v37 mom attend .702 v38 dad attend .686 Endogenous Latent Variables ACADEMIC ASPIRA./ v41 grade expect .609(**) .108(**) PERFORMANCE v43 grade receiv. .471 .859 v44 grade import. .098 .001(*) RELIGIOSITY v29 religious .578(**) v32 close to God .245 v33 pray .536 v36 attend .756 CONFORMITY TO v67,v53 par party .020(**) .293(**) .169(**) TRADITIONAL S.O. v49,v65 par date .321 .523 .431 v52,v58 chur lead .282 .119 .507 CONFORMITY TO v56 prof. letter .956(**) .256(**) RATION-LEGAL S.O. v62 prof. seat .040(*) .123 v68 prof. argue .049(*) .022(*) CONFORMITY TO v66,v54,v63 voting .216(**) .731(**) .467(**) MEDIA-PEER S.O. v60,v57,v51 buying .228 .211 .210 * Not significant at .05 probability level ** Reference variables for which no t values are calculated
Chi-square is usually based on the difference between the observed and the expected values, and the higher the chi-square, the more difference exists between the observed and the expected values. However, in a LISREL model, we hoped to find a small and nonsignificant chi-square because we expected a good theoretical model to fit the data. Therefore, a small chi-square relative to the degrees of freedom demonstrates the validity of the model. For these models the chi-squares are 124.39/104df for mainland China, 103.77/105df for Taiwan, and 517.19/242df for the U.S. With 1.196, .988, or 2.137 attending each degree of freedom, the chi-squares are considered to be small.
Root mean square residual is a measure of the average of the residual variances and covariances, and should be interpreted in relation to actual sizes of the observed variances and covariances. The root mean square residuals of these models are respectively .058, .064, .053. These figures indicate the average departure on each indicator of the model from the data. In other words, on average, about 6% unexplained variance attends each indicator in each sample. Therefore, the smaller the figure, the better.
The square multiple correlations for structural equations (see the numbers that go together with endogenous variables) indicate that the percentage of variance in the endogenous latent variables can be explained by the model.
The standardized solutions mean that an increase of one standard deviation in one independent variable will bring about the change in the dependent variable by that much of a standard deviation. Since the standardized coefficients can be compared with one another, the standardized solution instead of the maximum likelihood is used in the analyses of relations in the path models.
The curves with two directions connect the error terms for each pair of endogenous variables which are correlated with each other. The figure on each curve is the residual covariance, and the higher the figure, the higher the two unmeasured influences are correlated. The explanation of correlated errors is one of the most challenging areas for theorists if they choose to use LISREL as their analytical strategy (Thomas & Roghaar, 1990). The high values of residual covariance in the mainland China model indicate that many characteristics in one conformity variable are shared by the other two variables.
The structural models for mainland China and Taiwan are shown in Figure 2. As suggested by the hypothetical model, conformity to traditional significant others is strongly influenced by family integration: the more integrated the family, the more conformity of adolescents to their father, mother, and religion leader. Traditional conformity is also influenced by academic behavior--positively for mainland Chinese but negatively for Taiwanese. While Taiwanese females are more conforming to traditional significant others than are Taiwanese males, there is no obvious difference among the mainland Chinese youths.
Conformity to rational/legal significant others is influenced by gender in the mainland Chinese sample: females conform to the expectations of professors more than do males. But academic behavior is negatively related to rational/legal conformity in the Taiwan sample.
As to conformity to expectations of best friends and entertainment stars, the more integrated the family, the higher the mainland Chinese adolescents score. This direct relation was not found in the hypothetical model. As found in the hypothetical model, male Chinese are slightly more conforming to peer and entertainment stars' expectations. The Taiwan model shows only a positive relation between academic behavior and media/peer conformity.
In a comparison with the hypothetical model, the mainland China and Taiwan models fail to evidence effects of family SES, parents' religiosity, and respondents' own religiosity on conformity behavior.
In the U.S. sample, traditional conformity is negatively related to family SES, but positively related to family integration. However, the strongest predictor for American adolescents' conformity to traditional significant others is their religiosity level.
Rational/legal conformity is not predicted by any variable in this U.S. model. On the other hand, media/peer conformity is negatively correlated with adolescents' religiosity. Further, males are more likely to conform to the expectations of their best friends and entertainment stars.
While academic behavior does not have any effect on American adolescent conformity, religiosity is predicted by all the environmental factors.
Compared with the hypothetical model, religiosity was dropped from the two Chinese models while academic behavior was deleted from the American model. This consequence matches what we have found previously--that religion is a more important social institution for Americans than for the Chinese, while education is more important for the Chinese.
DISCUSSION AND CONCLUSION
With regard to conformity to parents' expectations, it was found that youths in the two Chinese societies scored higher than did American youths. The reason no difference was found as expected between mainland Chinese and Taiwanese requires further explanation. Although Taiwan society is much more modernized than mainland China, Taiwanese youths are still traditional in many ways, especially in the family. There are at least two possible explanations for this fact. First, Taiwan is more modernized than mainland China, but the modernization level in terms of economic development is much lower than in the U.S. Compared with other western societies, Taiwan achieved modernization without the conventional process experienced by others. Further, Taiwan has had a shorter history of industrialization and urbanization, and the traditional culture may not have changed very quickly. Second, the deeply rooted Confucianism, that emphasizes human relations and family cohesion, may not soon disappear as modernization comes to the Taiwan society. Without the influence of communist or Marxist ideology, Confucianism has remained one of the dominant teachings in Taiwan.
In conformity to religious figures, American youths scored the highest and the Chinese youths the lowest. The path model analyses also show that religion is not an important social institution in mainland China. This finding is in opposition to the modernization and secularization theory. Because of the communist and Marxist ideology imposed on mainland Chinese society in the past 40 years, it is easy to understand why there is not much variance in the religiosity variables (parents' religiosity and self religiosity) and why the two religiosity variables in the mainland China model did not predict adolescent conformity behavior and were deleted from the original model.
The lower religious participation and religiosity level for the Chinese people is not accounted for only by the chinese political structure. Even before the communist takeover of China in 1949, religion was considered by many people as superstition, which was practiced by career monks, nuns, or witches. In the history of China most people held negative attitudes toward religion and religious practices. This cultural tradition has been carried on by the Taiwanese. Religion is not as important as other social institutions in their society. This explains why the religiosity variables did not predict Taiwanese adolescent conformity well.
Cultural characteristics in the different societies also account for the situation in the U.S. The high level of individualism in American society may lead to the high percentage of religious Americans. Socialization activities are more limited to the nuclear family in societies where individualism and privacy are emphasized. In this type of society, the church provides frequent opportunities and an easily accessible place for people to meet their social needs. This point is further supported by the fact that less than 10% of regular church-goers in America go there for strictly religious reasons (Bilheimer, 1983).
Another explanation for the high percentage of religious people in the U.S. might lie in its history. Americans have a heritage of diverse religions. Many of the first colonists who came from various European countries several hundred years ago were pilgrims who had been persecuted by their rulers because of their religious beliefs. They were extremely religious, and their offspring are also very religious.
In conformity to educators' expectations, youths from the two Chinese societies scored higher than did American youths. This finding obviously goes against the modernization theory, but was predicted by the cultural determinism. Chinese people have respected teachers and have highly valued education from the beginning of its history. Confucianism again plays an important role in this anti-modernization theory finding. Influenced by both modernization and Confucianism, youths in Taiwan are supposed to score the highest in conforming to educators' expectations. Although the difference between mainland Chinese and Taiwanese is not statistically significant, the tendency is for Taiwanese students to conform more to their teachers than do the mainland Chinese.
In summary, contrary to modernization theory's predictions, the social institution of education is less important for American adolescents than for the Chinese. In addition, religion appears to be much more important in American society, which is also in opposition to modernization predictions. Further, explanations of the unexpected findings are more likely be found in the culture than anywhere else.
Modernization theory has received criticism in recent years from political scientists and political economists since it neglected cultural, historic, and social structural factors in its analysis (Chirot, 1986; Black, 1991; Wallerstein, 1980). Although little criticism has thus far been applied to modernization theory in the field of social psychology, this does not mean that it is free of modification. Results of the analyses of the current data support the argument that modernization theory predicting adolescent conformity behavior has to take into consideration the cultural differences. Future research must include cultural variables in any modernization formulations. Modernization theory will then undoubtedly be modified by cultural considerations.
1 The Europa World Year Book, 1991 and Statistical Abstract of the United States 1990 provide the following statistics: GNP per capita (USA: $18,570, Taiwan: $5,126, mainland China: $438), number of telephones per 100 population (U.S.: 76, Taiwan: 31, mainland China: 0.7), and population in agriculture (USA: 1.36%, Taiwan: 5.30%, mainland China: 30.72%).
2 Although the separation of Taiwan from mainland China in the past 40 years has brought about value differences between the two societies, the people share basically the same cultural tradition. Two thousand-year old Confucianism still dominates in the culture of both mainland China and Taiwan. Compared with the U.S. culture, the Chinese and Taiwanese have more similarities than differences.
Albrecht, S. L., Chadwick, B. A., & Alcorn, D. S. (1977). Religiosity and deviance. Journal for the Scientific Study of Religion, 16, 263-274.
Barry, H. III, Child, I. L., & Bacon, M. K. (1959). Relation of child training to subsistence economy. American Anthropologist, 61, 51-63.
Bilheimer, R. S. (Ed.) (1983). Faith and ferment: An interdisciplinary study of Christian beliefs and practices. Minneapolis, MN: Augsburg.
Black, J. K. (1991). Development in theory and practice: Bridging the gap. Boulder, CO: Westview Press.
Chirot, D. (1986). Social change in the modern era. New York: Harcourt Brace Jovanovich.
Chusmir, L. H., & Koberg, C. S. (1988). A look at sex differences in the relationships between religious beliefs and work-related attitudes. Journal of Social Behavior and Personality, 3, 37-48.
Coleman, J. S., Campbell, E. Q., Hobson, C. J., McPartland, J., Mood, A. M., Weinfeld, F. D., & York, R. L. (1966). Equality of educational opportunity. Washington, DC: U.S. Government Printing Office.
Collins, R. (1974). The empirical validity of the conflict tradition. Theory and Society, 1, 148-178.
Cooper, H. (1979). Statistically combining independent studies: A meta-analysis of sex differences in conformity research. Journal of Personality and Social Psychology, 37, 131-146.
Durkheim, E. (1897). Suicide (J. A. Spaulding, & G. Simpson, Trans. in 1951). New York: Free Press.
Eagly, A. H. (1978). Sex differences in influenceability. Psychological Bulletin, 85, 86-116.
Eagly, A. H., & Chrvala, C. (1986). Sex differences in conformity: Status and gender role interpretations. Psychology of Women Quarterly, 10, 203-220.
Inkeles, A (1972). Modernizing influences: Causes and consequences of individual change in six developing countries. Centro-Sociale, 19, 106-108.
Jencks, C., Smith, M., Acland, H., Bane, M. J., Cohen, D., Gintis, H., Hezns, B., & Michelson, S. (1972). Inequality: A reassessment of the effect of family and schooling in America. New York: Basic Books.
Kohn, L. (1959a). Social class and parental values. American Journal of Sociology, 64, 337-351.
Kohn,. L. (1959b). Social class and the experience of parental authority. American Sociological Review, 24, 352-366.
Kohn, M. L. (1963). Social class and parent-child relationship: An interpretation. American Journal of Sociology, 68, 471-480.
Kohn, M. L. (1976). Social class and parental values: Another confirmation of the relationship. American Sociological Review, 41, 538-545.
Lee, G. R. (1982). Family structure and interaction: A comparative analysis (2nd ed.). Minneapolis: University of Minnesota Press.
Lee, G. R. (1984). The utility of cross-cultural data: Potentials and limitations for family sociology. Journal of Family Issues, 5, 519-541.
Long, J. S. (1983). Confirmatory factor analysis. Series: Quantitative Application in the Social Sciences. Beverly Hills, CA: Sage Publications.
Maccoby, E., & Jacklin, C. (1974). The psychology of sex differences. Stanford, CA: Stanford University Press.
McGuire, W. J. (1985). Attitudes and attitude change. In G. Lindzey, & E. Aronson (Eds.), The Handbook of Social Psychology (3rd ed.). vol. 2. New York: Random House.
Milgram, S. (1961). Nationality and conformity. Scientific American, 205, 45-51.
Milgram, S. (1963). Behavioral study of obedience. Journal of Abnormal and Social Psychology, 67, 371-378.
Milgram, s. (1964). Group pressure and action against a person. Journal of Abnormal and Social Psychology, 69, 137-143.
Milgram, S. (1974). Obedience to authority. New York: Harper & Row.
Rigby, K., & Densley, T. R. (1985). Religiosity and attitude toward institutional authority among adolescents. Journal of Social Psychology, 125, 723-728.
Rollins, B.C., & Thomas, D. L. (1979). Parental support, power and control techniques in the socialization of children. In W. R. Burr, R, Hill, R. I. Nye, & I. L. Reiss (Eds.), Contemporary theories about the family (Vol. I). New York: Free Press.
Singh, S. N. (1984). The effect of socioeconomic status on social conformity. Perspectives in Psychological Researchers, 7, 36-37.
Smith, C. J. (1991). China: People and places in the land of one billion. San Francisco: Westview Press.
Stark, R. (1984). Religion and conformity: Reaffirming a sociology of religion. Sociological Analysis, 45, 273-282.
Thomas, D. L., & Weigert, A. J. (1971). Socialization and adolescent conformity to significant others: A cross-national analysis. American Sociological Review, 36, 835-847.
Thomas, D. L., & Weigert, A. J. (1972). Determining nonequivalent measurement in cross-cultural family research. Journal of Marriage and the Family, 34, 166-176.
Thomas, D. L., Franks, D. D., & Calonico, J. M. (1972). Role-taking and power in social psychology. American Sociological Review, 37, 605-614.
Thomas, D. L., & Roghaar, H. B. (1990). Postpositivit theorizing: The case of religion and the family. In J. Sprey (Ed.), Fashioning family theory: New approaches. Newbury Park: Sage Publications.
Thomas, D. L., Gecas, V., Weigert, A. & Rooney, E. (1974). Family socialization and the adolescent. Lexington, MA: Lexington Books.
Tuthill, D. M., & Forsyth, D. R. (1982), Sex differences in opinion conformity and dissent. Journal of Social Psychology, 116, 205-210.
Wallerstein, I. M. (1980). The modern world-system. New York: Academic Press.
Weber, M. (1922). Economy and society. New York: Bedminister.
Weigert, A. J., & Thomas, D. L, (1979). Family socialization and adolescent conformity and religiosity: An extension to Germany and Spain. Journal of Comparative Family Studies, 10, 371-383.
Darwin L. Thomas, Ph.D., Professor, Brigham Young University, Department of Sociology, Provo, UT 84602.…
Questia, a part of Gale, Cengage Learning. www.questia.com
Publication information: Article title: Modernization Theory Revisited: A Cross-Cultural Study of Adolescent Conformity to Significant Others in Mainland China, Taiwan, and the USA. Contributors: Jie, Zhang - Author, Thomas, Darwin L. - Author. Journal title: Adolescence. Volume: 29. Issue: 116 Publication date: Winter 1994. Page number: 885+. © 1999 Libra Publishers, Inc. COPYRIGHT 1994 Gale Group.
This material is protected by copyright and, with the exception of fair use, may not be further copied, distributed or transmitted in any form or by any means.