Evaluating Survey Items of Buddhism Religiosity in China

Evaluating Survey Items of Buddhism Religiosity in China Over the past few decades, public opinion studies have paid increasing attention to the comparison of eastern and western religions, an endeavor that calls for a rigorous evaluation of empirical measures of religiosity for eastern religions. This study investigates the applicability to Chinese Buddhism of a measurement scheme based on believing, belonging, and behaving (the Three Bs). Confirmatory factor analyses of the Chinese Spiritual Life Survey suggest that the Three Bs are valid measures of Buddhist religiosity as they converge to the same latent construct. They are also reliable survey instruments, as illustrated by measurement invariance across subpopulations defined by differences in region of residence, gender, cohort, and education. The Three Bs measurement scheme has potential for empirical comparative research. Evaluating Survey Items of Buddhism Religiosity in China It has been noted that religion is strongly related to individual attitudes toward political parties, abortion, gay rights, and foreign policy, to name a few (Baumgartner, Francia, & Morris, 2008; Clements, 2014a; Eisenstein, 2006). With the rise in cross-national comparative studies, various religions, especially ones from the East, have begun to enter the research agenda of empirical researchers (Bender, Cadge, Levitt, & Smilde, 2012; Cadge, Levitt, & Smilde 2011; Gonzalez, 2011; Tanaka, 2010). This brings the following questions to the fore: How can researchers measure religious involvement in different religions using survey instruments? More importantly, is it possible to gauge religiosity in different religions with a common measurement scheme?1 To examine this question, this research note presents a case study that explores whether a measurement scheme originally developed for Christianity can be extended to the eastern religion of Buddhism in China. We chose Buddhism as our subject not only because of its well-documented worldwide presence and social influences (McMahan, 2011) but also because of its notable tendency to growth in traditionally Christian nations (Leamaster, 2012). In light of this, a meticulous evaluation of survey items pertaining to Buddhist religiosity serves the interest of comparative public opinion studies. We choose China as our research site because the two major challenges to the reliability and validity of the empirical measurement of religiosity in the east are loose organization and theological syncretism, and these are typical in Chinese society (Leamaster & Hu, 2014; Hu, 2016). By virtue of this case study, we find evidence for the validity and reliability of a common religiosity measurement scheme for Buddhism in China, which strengthens our confidence in the applicability of similar survey items in other countries. The Three Bs Measurement Scheme In this study, we draw on a religiosity measurement scheme that centers on the Three Bs: believing, behaving, and belonging. “Believing” describes a follower’s faith and confidence in a religion; “behaving” describes the practices and activities that are driven by one’s religious commitment; “belonging” refers to affiliation and identification with a particular religious tradition (Green, Guth, Smidt, & Kellstedt, 1996; Olson & Warber, 2008). The Three Bs scheme has mainly been applied to investigate Christianity, and it has been shown that each of the Three Bs reflects the same underlying latent construct of Christian religiosity (Clayton, 1971; Clayton & Gladden, 1974; de Jong, Faulkner, & Warland, 1976; Faulkner & de Jong, 1966). In addition, the Three Bs are recognized as the most fundamental and stable dimensions of Christian religiosity (Glock and Stark 1965).2 As a result, survey instruments concerning these three dimensions of religiosity have often been used together to investigate various empirical themes, such as parenting value (Sieben & Halman, 2014), public service participation (Freeman & Houston, 2010), opposition to abortion (Clements, 2014b), public opinion about same-sex marriage (Olson, Cadge, & Harrison, 2006), and voting behavior (Smith & Walker, 2012). Despite the wide-ranging utilization of the Three Bs measurement scheme in Christianity, it is not clear whether it can be also applied to Buddhism in the East. From the Durkheimian functionalist perspective on religion, we speculate that the religiosity of Buddhists, like that of Christians, can also be gauged by the Three Bs,3 as the essential elements of religion are ritual practices (behaving) and the ensuing beliefs and communal belonging (Marshall 2002). Before moving on to the empirical section, we first familiarize readers with the Chinese context, by highlighting potential challenges to the validity and reliability of survey items concerning Buddhist religiosity. The Case of Chinese Buddhism Challenges to Measurement Validity Measurement validity refers to whether the variations of empirical items are driven by the same theoretical construct (Gliner, Morgan, & Leech, 2009). A valid measurement scheme for religiosity should guarantee that all the empirical items reflect the same latent construct of religious propensity. This, however, might not be the case for the Three Bs because of some unique features of Chinese Buddhism. First, some Chinese citizens’ Buddhist identification may not be fully driven by religiosity. As suggested by Lu and Li (2011), one possible reason for some Chinese citizens’ identification with Buddhism is that they are not familiar with religions other than Buddhism, so, for them, belonging to Buddhism becomes a routine cultural identity. Second, the religious implication of Buddhist practices might be vague in the minds of some practitioners, as these practices are syncretic, insofar as being practiced by followers of multiple religions. This overlap of Buddhism with other Chinese religions (e.g., Daoism and folk religion) has been highlighted by scholars with the term “folk Buddhism” (Lin, 2008; Overmyer, 1976). Third, the utilitarian orientation of Chinese Buddhism means that believing in Buddhism in China is not necessarily an indicator of a person’s transcendental commitment and faith, but may rather be simply a means of satisfying one’s pragmatic pleas (Chau 2006; Nadeau, & Komjathy, 2012).4 In summary, extant studies on Chinese Buddhism indicate that the measurement items for behaving, believing, and belonging, when applied to Buddhism, may entail factors that are too diverse to converge on the same latent construct of religiosity, challenging measurement validity. Challenges to Measurement Reliability In addition to validity, studies on Chinese Buddhism imply some potential concerns about the reliability of the Three Bs. Reliability, defined as the consistency of measurement items, calls for measurement invariance across subgroups of respondents.5 In China, this property is confronted with challenges. For instance, females might be less active than males in joining formal Buddhist organizations (Cadge, 2004). Instead, they are more likely to be unorganized lay Buddhists (Fan, 2003). Therefore, Buddhist religiosity measures may reveal a gender difference, with women attaching more significance to believing and behaving rather than to belonging. Besides gender, measurement inconsistencies could emerge between educational groups. The better educated pay more attention to self-cultivation and the ethical meaning of Buddhism, such that they value believing more than behaving. In contrast, the less educated are more superstitious, so for them, Buddhist involvement gravitates toward the dimension of behaving (Hu & Leamaster, 2013). A third sociodemographic factor that might challenge the Three Bs’ reliability is cohort. Specifically, the generations who experienced the socialist regime in their formative stage might be reluctant to openly display their religious practices, but rather express their religious commitment through private belief. This is partly because of the fact that the state, during the socialist regime, suppressed people’s religious beliefs less efficiently than it did their overt religious activities (Lee 2007; Leung, 2005; Potter, 2003). As a result, an intergenerational inconsistency could emerge as the post-Reform generations have been subject to state repression to a much lesser extent. Finally, the urban–rural difference may be another source of measurement inconsistency. The commercialization of urban Buddhism encourages disciplinary laxity among clergies, resulting in a hollowing-out of the “plausibility” of urban Buddhism (Sun, 2011); therefore, urban residents’ belief and faith in Buddhism, relative to their rural counterparts, could be less firm (Oakes & Sutton, 2010). Altogether, prior literature indicates that the Three Bs measurement scheme, when applied to Chinese Buddhism, meets potential challenges in both survey validity and survey reliability that question its applicability to cross-national comparative public opinion research. In this research note, we rigorously respond to these concerns, and empirically investigate the quality of the Three Bs survey items. Data, Measures, and Analytical Strategy Sample This study analyzes data from the Chinese Spiritual Life Survey (CSLS), which was collected in 2007. The CSLS was the first representative nationwide survey with a focus on Chinese residents’ religious life. The total number of respondents of the CSLS was 7,021, with an age interval between 16 and 75 years. The survey was administered in 56 locales throughout mainland China. Within each locale, neighborhoods were sampled within administratively defined neighborhood committees (government-defined collections of neighborhoods), and households were randomly sampled within each neighborhood. A KISH grid procedure was used to randomly select one respondent for a face-to-face interview. The final sample had an American Association for Public Opinion Research (AAPOR) response rate of 28.1%.6 Although the CSLS was collected in 2007, it is to date the only survey that specifically focuses on the various aspects of Chinese citizens’ religious lives. Other than the CSLS, no other surveys (e.g., the Chinese General Social Survey conducted in 2010) provide all of the survey instruments of the Three Bs. Since being collected, the CSLS has been used by scholars to study a wide range of topics pertaining to Chinese religions (Liu and Mencken, 2010, Stark and Liu, 2011, Yang and Hu, 2012, and Zhai, 2010). Measures Both the Chinese and English wordings of the items are presented in Table A1. To shed light on comparative studies, we selected the most common measurement items for Buddhist religiosity that are applicable, as much as possible, to the various schools of Buddhism. Because of this, some survey items, which are more school-specific, were not used.7 A common practice in studies of Christianity is to measure belonging with denominational affiliation (e.g., Baptist or Methodist). However, no such nuanced denominational distinctions apply in Chinese Buddhism. Moreover, even if we distinguish between schools of Buddhism, a person simply considering himself/herself affiliated with one particular Buddhist tradition (e.g., the Mahayana) is insufficient in revealing his/her organizational attachment because of the syncretic nature of religious life and the vague meaning of religious affiliation in the Chinese context (Leamaster & Hu, 2014). Because of this, we argue that an item that can better gauge one’s “belonging” to Buddhism would be a stricter indicator of organizational involvement. Specifically, we measure Buddhist “belonging” using the question of whether a respondent converted to Buddhism through a formal guiyi ritual (conversion) (1 = yes; 0 = no).8 According to the theology of Buddhism, those who go through the guiyi ritual are considered to have taken refuge in the Three Jewels (Buddha, Dharma, and Sangha) (Kurtz, 2007; Welch, 1972). As noted by Hu and Leamaster (2013), the guiyi ritual implies an extensive and somewhat exclusivist commitment to Buddhism, so it should serve better than self-reported affiliation as a measure of one’s organizational belonging to Buddhism.9 The dimension of “believing” concerns people’s faith and confidence in Buddhism; this is measured by two questions (1 = yes; 0 = no). The first question directly asks whether one believes in Buddhism (belief) from a list of religions (Buddhism, Catholicism, Protestantism, Taoism, Islam, and others).10 Note that, in this study, we view this variable as an item relevant to the dimension of “believing” instead of “belonging” because the wording of this question gravitates more toward belief in (xin) Buddhism rather than organizational religious affiliation. As an individual’s attitudes toward the existence of spiritual figures in a religion represent a fundamental aspect of religious faith (Polkinghorne, 1998), our second measure for “believing” is a question about whether one believes that Buddha (fozu) and Bodhisattva (pusa) exist (existence) (1 = yes; 0 = no). It is worth mentioning that this question measures people’s Buddhist belief instead of Chinese folk religious belief because (1) the wording of the questionnaire made it clear that the subjects in question are fozu and pusa, two Buddhist figures; and (2) folk religious deities and spirits had already been listed elsewhere in the CSLS.11 Buddhist practices in China are diverse, such that it is almost impossible to exhaustively consider all of them. In this study, we chose some typical practices considered in the CSLS to measure “behaving” in Buddhism. Specifically, three questions were used, including whether participants burnt incense for Buddha and Bodhisattva in a Buddhist temple (temple), whether respondents paid respects to Buddha and Bodhisattva (worship), and whether they read or study Buddhist scriptures (sutra) (1 = yes; 0 = no). These questions represent the most common Buddhist practices in China, involving both private and public, as well as both organizational and nonorganizational activities (Birnbaum, 2003).12 The CSLS is, to date, the only large-scale survey that focuses exclusively on Chinese citizens’ religious life, but there is still room for further improvement. For example, the CSLS does not provide information about the more subtle variations between different traditions of Chinese Buddhism (e.g., the Vajrayana school of Buddhism) and other Buddhism-inspired practices. Hence, the items used here capture Buddhist involvement in the general sense. Also, as a study based on secondary data, we were not able to design our own items covering everything related to Buddhist involvement, so it is inevitable that the examined items are relatively simple in both form and content (e.g., several binary items). Finally, the self-presentation effect and social-expectation bias cannot be fully ruled out. For example, people who do not claim Buddhist “belonging,” “behaving,” or “believing” may still be involved in them in private.13 Notwithstanding these limitations, however, we believe that a formal evaluative study on the performance of the measurement instruments of the Three Bs for the case of Buddhism, one major religion in the East, is still meaningful, necessary, and timely in the light of increasing attention toward eastern religions in public opinion research. Analytical Strategy As stated earlier, empirical items are valid if they quantify the same underlying theoretical construct. To assess the validity of the Three Bs, we performed a confirmatory factor analysis (CFA) in which a single latent variable of Buddhist religiosity was configured for all of the observed items. In addition to this model, we also configured a CFA model in which multiple latent constructs were configured to, respectively, correspond to religious belonging, believing, and behaving.14 This more complicated model showed good model fit, but it was not significantly better than the one with a single latent variable.15 Because of this, we reported the CFA model in which all items were mapped to the single latent construct. Reliability was examined by considering the measurement invariance of the Three Bs items across sociodemographic groups, which was performed through multigroup confirmatory factor analysis (MGCFA). In MGCFA, there are three levels of measurement invariance (Meredith, 1993). The basic level is configural invariance, meaning that one model configuration can fit the data of different groups and that all item loadings are statistically significant. The second level is metric invariance. This level constrains factor loadings to be identical across groups. The third level is scalar invariance, where both factor loadings and intercepts are constrained as invariant across groups. Scalar invariance is stricter than metric invariance, which is, in turn, stricter than configural invariance. In the following research, we tested the reliability of the Three Bs measurement items for Chinese Buddhism across subpopulations defined by gender (male vs. female), education (below high school vs. high school and above), region of residence (urban vs. rural), and cohort (pre-1978 cohort vs. post-1978 cohort).16 In the following analyses, multiple indexes of model fit were used, including the Chi-square value, the root mean squared error of approximation (RMSEA), the comparative fit index (CFI), and the Tucker–Lewis index (TLI). RMSEA values <.05 indicate excellent fit (Hu and Bentler, 1999). The range of CFI and TLI is from 0 (poor fit) to 1 (perfect fit), and good model fit should return a value >.95.17 Results Descriptive Patterns Descriptive patterns pertaining to Chinese Buddhism can be found in Table 1. As shown, around 2% of the respondents of the CSLS had formally converted to Buddhism through the ritual of guiyi. In total, 7% believed that Buddha and Bodhisattva exist, and 17% reported belief in the religion of Buddhism. In terms of Buddhist practices, 11% of Chinese adults had burnt incense for Buddha and Bodhisattva in a Buddhist temple; comparatively, fewer individuals had paid respect to Buddha and Bodhisattva (6%) or read or studied Buddhist scriptures (1%). Table 1 Descriptive Statistics Variable Description % SD Conversion Formally convert to Buddhism through the guiyi ceremony 2 0.12 Existence Believing that Buddha and Bodhisattva exist 7 0.26 Belief Believing in Buddhism 17 0.37 Temple Burning incense for Buddha and Bodhisattva in a Buddhist temple 11 0.31 Worship Paying respect to Buddha and Bodhisattva 6 0.23 Sutra Reading/studying Buddhist scriptures 1 0.1 N 7,021 Variable Description % SD Conversion Formally convert to Buddhism through the guiyi ceremony 2 0.12 Existence Believing that Buddha and Bodhisattva exist 7 0.26 Belief Believing in Buddhism 17 0.37 Temple Burning incense for Buddha and Bodhisattva in a Buddhist temple 11 0.31 Worship Paying respect to Buddha and Bodhisattva 6 0.23 Sutra Reading/studying Buddhist scriptures 1 0.1 N 7,021 Source: Chinese Spiritual Life Survey 2007. Table 1 Descriptive Statistics Variable Description % SD Conversion Formally convert to Buddhism through the guiyi ceremony 2 0.12 Existence Believing that Buddha and Bodhisattva exist 7 0.26 Belief Believing in Buddhism 17 0.37 Temple Burning incense for Buddha and Bodhisattva in a Buddhist temple 11 0.31 Worship Paying respect to Buddha and Bodhisattva 6 0.23 Sutra Reading/studying Buddhist scriptures 1 0.1 N 7,021 Variable Description % SD Conversion Formally convert to Buddhism through the guiyi ceremony 2 0.12 Existence Believing that Buddha and Bodhisattva exist 7 0.26 Belief Believing in Buddhism 17 0.37 Temple Burning incense for Buddha and Bodhisattva in a Buddhist temple 11 0.31 Worship Paying respect to Buddha and Bodhisattva 6 0.23 Sutra Reading/studying Buddhist scriptures 1 0.1 N 7,021 Source: Chinese Spiritual Life Survey 2007. Evaluating Measurement Validity The results of the CFA model are presented in Figure 1. All of the six items are significantly associated with the same latent construct of Buddhist religiosity (p < .05). Among those items, specifically, the factor with the highest loading is belief (3.031), which is followed by worship (1.508), temple (1.457), existence (1.361), sutra (1.290), and conversion (1.181). Figure 1 View largeDownload slide Result of confirmatory factor analysis. Note: All factor loadings are statistically significant at the 0.001 level. CI = confidence interval; Rel. = religiosity. Source: Chinese Spiritual Life Survey 2007 Figure 1 View largeDownload slide Result of confirmatory factor analysis. Note: All factor loadings are statistically significant at the 0.001 level. CI = confidence interval; Rel. = religiosity. Source: Chinese Spiritual Life Survey 2007 To evaluate this model's goodness of fit, a series of indexes is reported. Specifically, the Chi-square value of this CFA model is 43.695. Given the 9 degrees of freedom, this value is significant at the .0001 level. This result supports the goodness of fit of this model. In addition to the Chi-square value, CFI (.997), TLI (.995), and RMSEA (.023, with 90% confidence interval to be .017–.031) all suggest excellent model fit. In summary, a CFA model with all Three Bs instruments directed at the same latent construct of Buddhist religiosity fits the CSLS data well, so the validity of the Three Bs is confirmed.18 Evaluating Measurement Reliability The results of the MGCFA can be found in Table 2. One way to check the reliability of the Three Bs is to investigate the model fit indexes across the three levels of measurement invariance. If the CFA continues to fit the data as we move from configural invariance to metric invariance and then to scalar invariance, we have evidence of measurement consistency across subpopulations. Table 2 Comparison of Nested Confirmatory Factor Analytic Models for Evaluating Invariance by Demographic Variables for the Measures of Buddhism Involvement Model # Free parameters Chi-square value Chi-square DF Chi-square p-value CFI ΔCFI TLI RMSEA estimate RMSEA 90% CI lower bound RMSEA 90% CI higher bound RMSEA p-value Region of residence     1. Configural model 24 61.604 18 .0000 .996 .994 .026 .019 .034 1.000     2. Metric model 18 59.461 24 .0001 .997 .001 .996 .021 .014 .027 1.000     3. Scalar model 12 81.111 30 .0000 .996 .000 .996 .022 .016 .028 1.000 Education     1. Configural model 24 58.513 18 .0000 .997 .994 .025 .018 .033 1.000     2. Metric model 18 57.036 24 .0002 .997 .000 .996 .020 .013 .026 1.000     3. Scalar model 12 67.128 30 .0001 .997 .000 .997 .019 .013 .025 1.000 Gender     1. Configural model 24 54.206 18 .0000 .997 .995 .024 .017 .031 1.000     2. Metric model 18 48.397 24 .0023 .998 .001 .997 .017 .010 .024 1.000     3. Scalar model 12 193.58 30 .0000 .985 −.013 .985 .039 .034 .045 .999 Cohort     1. Configural model 23 61.337 19 .0000 .996 .994 .025 .018 .032 1.000     2. Metric model 18 59.109 24 .0000 .997 .001 .996 .020 .014 .027 1.000     3. Scalar model 12 6.578 30 .0000 .997 .000 .997 .017 .011 .023 1.000 Model # Free parameters Chi-square value Chi-square DF Chi-square p-value CFI ΔCFI TLI RMSEA estimate RMSEA 90% CI lower bound RMSEA 90% CI higher bound RMSEA p-value Region of residence     1. Configural model 24 61.604 18 .0000 .996 .994 .026 .019 .034 1.000     2. Metric model 18 59.461 24 .0001 .997 .001 .996 .021 .014 .027 1.000     3. Scalar model 12 81.111 30 .0000 .996 .000 .996 .022 .016 .028 1.000 Education     1. Configural model 24 58.513 18 .0000 .997 .994 .025 .018 .033 1.000     2. Metric model 18 57.036 24 .0002 .997 .000 .996 .020 .013 .026 1.000     3. Scalar model 12 67.128 30 .0001 .997 .000 .997 .019 .013 .025 1.000 Gender     1. Configural model 24 54.206 18 .0000 .997 .995 .024 .017 .031 1.000     2. Metric model 18 48.397 24 .0023 .998 .001 .997 .017 .010 .024 1.000     3. Scalar model 12 193.58 30 .0000 .985 −.013 .985 .039 .034 .045 .999 Cohort     1. Configural model 23 61.337 19 .0000 .996 .994 .025 .018 .032 1.000     2. Metric model 18 59.109 24 .0000 .997 .001 .996 .020 .014 .027 1.000     3. Scalar model 12 6.578 30 .0000 .997 .000 .997 .017 .011 .023 1.000 Note. CFI = comparative fit index; CI = confidence interval; RMSEA = root mean squared error of approximation TLI = Tucker–Lewis index. Source: Chinese Spiritual Life Survey 2007. Table 2 Comparison of Nested Confirmatory Factor Analytic Models for Evaluating Invariance by Demographic Variables for the Measures of Buddhism Involvement Model # Free parameters Chi-square value Chi-square DF Chi-square p-value CFI ΔCFI TLI RMSEA estimate RMSEA 90% CI lower bound RMSEA 90% CI higher bound RMSEA p-value Region of residence     1. Configural model 24 61.604 18 .0000 .996 .994 .026 .019 .034 1.000     2. Metric model 18 59.461 24 .0001 .997 .001 .996 .021 .014 .027 1.000     3. Scalar model 12 81.111 30 .0000 .996 .000 .996 .022 .016 .028 1.000 Education     1. Configural model 24 58.513 18 .0000 .997 .994 .025 .018 .033 1.000     2. Metric model 18 57.036 24 .0002 .997 .000 .996 .020 .013 .026 1.000     3. Scalar model 12 67.128 30 .0001 .997 .000 .997 .019 .013 .025 1.000 Gender     1. Configural model 24 54.206 18 .0000 .997 .995 .024 .017 .031 1.000     2. Metric model 18 48.397 24 .0023 .998 .001 .997 .017 .010 .024 1.000     3. Scalar model 12 193.58 30 .0000 .985 −.013 .985 .039 .034 .045 .999 Cohort     1. Configural model 23 61.337 19 .0000 .996 .994 .025 .018 .032 1.000     2. Metric model 18 59.109 24 .0000 .997 .001 .996 .020 .014 .027 1.000     3. Scalar model 12 6.578 30 .0000 .997 .000 .997 .017 .011 .023 1.000 Model # Free parameters Chi-square value Chi-square DF Chi-square p-value CFI ΔCFI TLI RMSEA estimate RMSEA 90% CI lower bound RMSEA 90% CI higher bound RMSEA p-value Region of residence     1. Configural model 24 61.604 18 .0000 .996 .994 .026 .019 .034 1.000     2. Metric model 18 59.461 24 .0001 .997 .001 .996 .021 .014 .027 1.000     3. Scalar model 12 81.111 30 .0000 .996 .000 .996 .022 .016 .028 1.000 Education     1. Configural model 24 58.513 18 .0000 .997 .994 .025 .018 .033 1.000     2. Metric model 18 57.036 24 .0002 .997 .000 .996 .020 .013 .026 1.000     3. Scalar model 12 67.128 30 .0001 .997 .000 .997 .019 .013 .025 1.000 Gender     1. Configural model 24 54.206 18 .0000 .997 .995 .024 .017 .031 1.000     2. Metric model 18 48.397 24 .0023 .998 .001 .997 .017 .010 .024 1.000     3. Scalar model 12 193.58 30 .0000 .985 −.013 .985 .039 .034 .045 .999 Cohort     1. Configural model 23 61.337 19 .0000 .996 .994 .025 .018 .032 1.000     2. Metric model 18 59.109 24 .0000 .997 .001 .996 .020 .014 .027 1.000     3. Scalar model 12 6.578 30 .0000 .997 .000 .997 .017 .011 .023 1.000 Note. CFI = comparative fit index; CI = confidence interval; RMSEA = root mean squared error of approximation TLI = Tucker–Lewis index. Source: Chinese Spiritual Life Survey 2007. Based on the analytical results presented in Table 2, measurement equivalence is supported for variations in region of residence. RMSEA (<.05), CFI (>.95) and TLI (>.95) all suggest an excellent model fit at all three levels of measurement invariance. This finding means that people living in rural and urban areas in contemporary China perceive the relation between observed measurement items and latent Buddhist religiosity consistently. Similar results are obtained for group variations in gender, education, and cohort, as all three levels of measurement invariance are met according to the model fit indices (RMSEA <.05, CFI >.95, and TLI >.95). Taken together, this means that a consistent empirical link between the Three Bs items and latent Buddhist religiosity will be produced, whether the Three Bs measurement scheme is applied to females or males, people with different educational attainments, or people of different generations.19 Concluding Remarks Religion has been cited by public opinion researchers to account for people’s attitudes toward a variety of social issues. Over the past few decades, cross-national comparative studies have been increasing, directing public opinion researchers’ attention to various religions in different parts of the world, especially the East. However, both the validity and reliability of survey instruments for religiosity are threatened by the peculiar national sociocultural context, necessitating a formal evaluation of these measurement schemes. In this research note, we have presented a case study of Buddhism in China. Empirical results suggest that a measurement scheme based on religious belonging, behaving, and believing is valid and able to gauge the different aspects of a single latent construct of Buddhist religiosity. The reliability of this scheme is empirically confirmed by the measurement equivalence across regional, gender, generational, and educational subpopulations. That is to say, people of different sociodemographic backgrounds in China view the link between the observed items and latent Buddhist religiosity consistently. These findings suggest that, even in a country like China where people’s religious life is loosely organized and syncretic, the survey items of the Three Bs, which have mainly been applied to Christianity, demonstrate the same underlying indicator of religiosity and return reliable information across sociodemographic subgroups. In light of this, we expect the Three Bs measurement scheme to be applicable in other nations. For example, in the United States, followers of Buddhism have been well organized. Even those from the Zen tradition form “a dharma of place” to enhance collective belonging. In addition, Buddhist beliefs have been a source of solace for many people, and Buddhist ritualistic practices, such as the “empowerment” from a mentor, the way to show respect and veneration for the bodhi tree or animistic deities, have also been prevalent (Coleman, 2001; Storhoff & Whalen-Bridge, 2010; Seager, 1999). In this regard, embodiments of all Three Bs are abundant, which set the stage for the application of our examined measurement scheme beyond the East. In fact, we speculate that the survey validity and reliability of the Three Bs items would be even better than what we document here because Buddhist practices, beliefs, and affiliation in the United States, can be less syncretic than those in China. This study is encouraging in showing that the religiosity of Buddhists in an Eastern society is similar to that of Christians, in that it can be empirically gauged by the same three components of religiosity. Thus, future cross-national and cross-religion comparative studies can be conducted in a more confident and more subtle fashion. For example, the effects on public opinion of different religions (e.g., Christianity vs. Buddhism) can be elaborated upon in terms of not only the general interreligion variations but also the precise component variations of the Three Bs. Clearly, new insights can be gleaned from this type of research. Anning Hu is Professor of Sociology at Fudan University, Shanghai, China. His research interests include the Sociology of Religion, Culture, Trust, and Social Research Methodology. This study was supported by the School of Social Development and Public Policy of Fudan University, the shuguang scholar program (17SG08), and the chuangxin group project of Fudan University (IDH3458007). Footnotes 1The term “religiosity” denotes a latent and multidimensional construct that describes how religious a person is. Following the social scientific research paradigm, this construct should be empirically measured by a battery of items. 2Specifically, the ideological dimension captures people's belief and confidence in a religion. The intellectual aspect of religiosity refers to one’s knowledge of religious doctrines. The ritualistic dimension focuses on overt religious behaviors. Religious experience is defined as the feelings and emotions attached to one’s religion. Finally, the consequential dimension refers to a religion’s various effects on people's secular life. 3According to Durkheim (1912), different religions, notwithstanding their distinct theologies and literatures, are defined in terms of the common functions of fostering social integration and a sense of unity. 4This could be the case for Chinese Buddhism, especially in light of the purported magical efficacy (lingyan) of many popular Buddhist deities (Lopez, 1996). For example, Bodhisattva (guanyin) is one of the most popular Buddhist deities in Chinese culture. It is widely believed that guanyin grants most blessings that are requested. 5This is shown through the coefficient invariance in the following confirmatory factor analysis. Measurement invariance does not mean that different people have the same values for the observed items. 6This number is somewhat deflated because 18% of all households sampled were unoccupied/empty when the interviewer visited the address. To ensure data quality, a postsurvey team was formed to double check over 20% of the interviews via telephone. 7For example, in an earlier version of this research note, we included the item of “Prayer.” However, as pointed out by the editor, this measure can be called into question for the Vajrayana schools of Buddhism in China, where the focus is on meditation and experience. Although we have good reasons to suspect that the type of Buddhism in the survey areas of the CSLS is mostly Chinese Buddhism, we believe a general measurement scheme better serves the empirical comparative research. For the empirical analysis, excluding the item of “Prayer” improved the model fit of the confirmatory factor analysis. 8Guiyi refers to a ritual process through which a Buddhist adherent establishes a mentor–disciple relationship with a Buddhist master. Usually, before guiyi is performed, a follower has to learn Buddhism informally with the master for some time. After conversion through guiyi, a Buddhist often lives in a temple and becomes intensively involved in the collective ceremonies and classes there. These Buddhists must also maintain a strict monastic lifestyle. 9In this regard, Buddhist guiyi looks similar to Christian baptism, but entails the additional expectation of a mentor–disciple tie. This partly accounts for the relatively low percentage of formally converted Buddhists in the CSLS. 10One potential concern related to this measure is the culture of polytheism in China (Gries, Su, & Schak, 2012). Nevertheless, this concern should not be severe because most respondents only chose one option, although multiple options were allowed. 11This question was designed to emphasize the spiritual aspect of Chinese Buddhism. Respondents should not misunderstand Buddha to be Shakyamuni, a historical figure. 12A potential concern about these measures is that they might be driven by the general Chinese cultural tradition instead of Buddhist religiosity. An empirical test of this concern is one objective of this study. If this concern proves significant, the Three Bs measurement scheme will reveal weak validity. 13In this regard, a survey like the CSLS might underestimate the popularity of Buddhism as well as its various dimensions in contemporary China. 14The diagram of this model can be found in Figure A1. 15The values of CFI, TLI, and RMSEA for the two models are close. Moreover, the difference in the Chi-square value is 5.64 (degree of freedom =2), which is not statistically significant. 16We chose 1978 as the demarcation point because the comprehensive Reform was initiated in that year. 17To accommodate the binary nature of the observed instruments, we used the mean and variance-adjusted weighted least squares (WLSMV) estimator in Mplus (Muthén & Muthén, 2007). The WLSMV has the merit of being free from the normality assumption for observed items, so it is always viewed to be the best option for research using categorical items (Brown, 2006). 18In Figure A1, we also present the goodness of fit of the CFA model with multiple latent variables, which is also well fitted, but not necessarily better than the model reported here. 19For the sake of cross-validation, we also present the results of the MGCFA with multiple latent constructs in Table A2. Again, measurement invariance is affirmed. References Baumgartner J. C. , Francia P. L. , Morris J. S. ( 2008 ). A clash of civilizations? The influence of religion on public opinion of US foreign policy in the Middle East . Political Research Quarterly , 61 , 171 – 179 . Google Scholar CrossRef Search ADS Bender C. , Cadge W. , Levitt P. , Smilde D. ( 2012 ). Religion on the edge: De-centering and re-centering the sociology of religion . New York, NY : Oxford University Press . Google Scholar CrossRef Search ADS Birnbaum R. ( 2003 ). Buddhist China at the century's turn . The China Quarterly , 174 , 428 – 450 . Google Scholar CrossRef Search ADS Brown T. ( 2006 ). Confirmatory factor analysis for applied research . New York, NY : Guildford . Cadge W. ( 2004 ). Gendered religious organizations: The case of theravada buddhism in America . Gender and Society , 18 , 777 – 793 . Google Scholar CrossRef Search ADS Cadge W. , Levitt P. , Smilde D. ( 2011 ). De-centering and re-centering: rethinking concepts and methods in the sociological study of religion . Journal for the Scientific Study of Religion , 50 , 437 – 449 . Google Scholar CrossRef Search ADS Chau A. Y. ( 2006 ). Miraculous response: Doing folk religion in contemporary China . Stanford, CA : Stanford University Press . Clayton R. R. , Gladden J. ( 1974 ). The five dimensions of religiosity: Toward demythologizing a sacred artifact . Journal for the Scientific Study of Religion , 13 , 135 – 143 . Google Scholar CrossRef Search ADS Clayton R. R. ( 1971 ). 5-D or 1? Journal for the Scientific Study of Religion , 10 , 37 – 40 . Google Scholar CrossRef Search ADS Clements B. ( 2014a ). Religion and public opinion in Britain . New York, NY : Palgrave Macmillan . Clements B. ( 2014b ). Religion and the sources of public opposition to abortion in Britain: The role of ‘belonging’, ‘behaving’ and ‘believing.’ Sociology , 48 , 369 – 386 . Google Scholar CrossRef Search ADS Coleman J. W. ( 2001 ). The New Buddhism: The western transformation of an ancient tradition . New York, NY : Oxford University Press . De Jong G. F. , Faulkner J. , Warland R. ( 1976 ). Dimensions of religiosity reconsidered; evidence from a cross-cultural study . Social Forces , 54 , 866 – 889 . Google Scholar CrossRef Search ADS Durkheim E. ( 1912 ). The elementary forms of religious life . Thousand Oak, CA : Sage Publications . Eisenstein M. ( 2006 ). Rethinking the relationship between religion and political tolerance in the US . Political Behavior , 28 : 327 – 348 . Google Scholar CrossRef Search ADS Emerson M. O. , Mirola W. , Monahan S. ( 2010 ). Religion matters: What sociology teaches us about religion in our world? New York, NY : Pearson . Emmons R. A. , Paloutzian R. ( 2003 ). The psychology of religion . Annual Review of Psychology , 54 , 377 – 402 . Google Scholar CrossRef Search ADS PubMed Fan H. ( 1997 ). Footbinding, feminism and freedom: The liberation of women's bodies in modern China . New York, NY : Routledge . Fan L. ( 2003 ). The cult of the silkworm mother as a core of local community religion in a North China Village: Field study in Zhiwuying, Baoding, Hebei. In Overmyer D. (Ed.), Religion in China today (pp. 53 – 66 ). New York, NY : Cambridge University Press . Faulkner J. E. , de Jong G. ( 1966 ). Religiosity in 5-D: An empirical analysis . Social Forces , 45 , 246 – 254 . Google Scholar CrossRef Search ADS Freeman P. K. , Houston D. J. ( 2010 ). Belonging, believing, and behaving: The religious character of public servants . Administration and Society , 42 , 694 – 719 . Google Scholar CrossRef Search ADS Froese P. ( 2008 ). The plot to kill god: Findings from the soviet experiment in secularization . Berkeley, CA : University of California Press . Google Scholar CrossRef Search ADS Fukuyama Y. ( 1961 ). The major dimensions of church membership . Review of Religious Research , 2 , 154 – 161 . Google Scholar CrossRef Search ADS Gliner J. A. , Morgan G. , Leech N. ( 2009 ). Research design and analysis in applied settings: An integrated approach to design and analysis . New York, NY : Routledge . Glock C. Y. , Stark R. ( 1965 ). Christian beliefs and anti-semitism . New York, NY : Harper and Row . Gonzalez A. ( 2011 ). Measuring religiosity in a majority Muslim context: Gender, religious salience, and religious experience among Kuwaiti college students—a research note . Journal for the Scientific Study of Religion , 50 , 339 – 350 . Google Scholar CrossRef Search ADS Green J. , Guth J. , Smidt C. , Kellstedt L. , ( 1996 ). Religion and the culture wars: Dispatches from the front . Lanham, MD : Rowman & Littlefield Publishers . Gries P. , Su J. , Schak D. ( 2012 ). Toward the scientific study of polytheism: Beyond forced‐choice measures of religious belief . Journal for the Scientific Study of Religion , 51 , 623 – 637 . Google Scholar CrossRef Search ADS Hu A. , Leamaster R. ( 2013 ). Longitudinal trends of religious. Groups in deregulated Taiwan: 1990-2009 . Sociological Quarterly , 54 , 254 – 277 . Google Scholar CrossRef Search ADS Hu A. ( 2016 ). Ancestor worship in contemporary China: An empirical investigation . China Review: An Interdisciplinary Journal on Greater China , 16 , 169 – 186 . Hu L. , Bentler P. ( 1999 ). Cutoff criteria for fit indexes in covariance structure analysis: Conventional criteria versus new alternatives . Structural Equation Modeling , 6 , 1 – 55 . Google Scholar CrossRef Search ADS Kurtz L. ( 2007 ). Gods in the global village: The world's religions in sociological perspective . Thousand Oaks, CA : Sage Publications . Leamaster R. , Hu A. ( 2014 ). Popular Buddhists: Predicting popular religious belief and practice in contemporary China . Sociology of Religion , 75 , 234 – 259 . Google Scholar CrossRef Search ADS Leamaster R. J. ( 2012 ). A research note on english‐speaking Buddhists in the United States . Journal for the Scientific Study of Religion , 51 , 143 – 155 . Google Scholar CrossRef Search ADS Lee T. ( 2007 ). Christianity in contemporary China: An update . Journal of Church and State , 49 , 277 – 304 . Google Scholar CrossRef Search ADS Leung B. ( 2005 ). China’s religious freedom policy: The art of managing religious activity . The China Quarterly , 184 , 894 – 913 . Google Scholar CrossRef Search ADS Lin M. ( 2008 ). Mazu Xinyang yu Hanren Shehui [Belief of Mazu and Han Chinese Society in Taiwan] . Ha-er-bing, China : Helongjiang Renmin Chubanshe . Liu E. , Mencken C. ( 2010 ). Fatalistic voluntarism and life happiness in post-socialist China . Sociological Spectrum , 30 , 270 – 288 . Google Scholar CrossRef Search ADS Lopez D. ( 1996 ). Religions of China in practice . Princeton, NJ : Princeton University Press . Lu Y. , Li D. ( 2011 ). The pattern, contemporary situation, and trends of Taiwan Religions. In Ji Z. , Qiu Y. (Eds.) The annual report on religions in China (pp. 225 – 252 ). Beijing, China : Chinese Social Sciences Press . Marshall D. ( 2002 ). Behavior, belonging, and belief: A theory of ritual practice . Sociological Theory , 20 , 360 – 380 . Google Scholar CrossRef Search ADS McMahan D. L. ( 2011 ). Buddhism in the Modern World . New York, NY : Routledge . Meredith W. ( 1993 ). Measurement invariance, factor analysis, and factorial invariance . Psychometrika , 58 , 525 – 543 . Google Scholar CrossRef Search ADS Muthén L. , Muthén B. ( 2007 ). Mplus user’s guide 4 . Los Angeles, CA : Author . Nadeau R. , Komjathy L. ( 2012 ). The wiley-blackwell companion to Chinese religions . New York, NY : Wiley-Blackwell . Google Scholar CrossRef Search ADS Oakes T. , Sutton D. ( 2010 ). Faiths on display: Religion, tourism, and the Chinese state . Lanham, MD : Rowman & Littlefield Publishers . Olson L. , Warber A. ( 2008 ). Belonging, behaving, and believing: Assessing the role of religion on presidential approval . Political Research Quarterly , 61 , 192 – 204 . Google Scholar CrossRef Search ADS Olson L. R. , Cadge W. , Harrison J. T. ( 2006 ). Religion and public opinion about same‐sex marriage . Social Science Quarterly , 87 , 340 – 360 . Google Scholar CrossRef Search ADS Overmyer D. ( 1976 ). Folk Buddhist religion: Dissenting sects in late traditional China . Cambridge, MA : Harvard University Press . Google Scholar CrossRef Search ADS Polkinghorne J. ( 1998 ). Belief in god in an age of science . New Haven, CT : Yale University Press . Potter P. ( 2003 ). Belief in control: Regulation of religion in China . The China Quarterly , 174 , 317 – 337 . Seager R. H. ( 1999 ). Buddhism in America . New York, NY : Columbia University Press . Sieben I. , Halman L. ( 2014 ). Religion and parental values in a secularized country: Evidence from the Netherlands . Social Compass , 61 , 121 – 140 . Google Scholar CrossRef Search ADS Smith L. , Walker L. ( 2012 ). Belonging, believing, and group behavior: Religiosity and voting in American Presidential Elections . Political Research Quarterly , 66 , 399 – 413 . Google Scholar CrossRef Search ADS Stark R. , Glock C. ( 1968 ). Patterns of religious commitment . Berkeley, CA : University of California Press . Stark R. , Liu E. ( 2011 ). The religious awakening in China . Review of Religious Research , 52 , 282 – 289 . Storhoff G. , Whalen-Bridge J. ( 2010 ). American Buddhism as a Way of Life . New York, NY : SUNY Press . Sun Y. ( 2011 ). The Chinese Buddhist ecology in post-mao China: Contours, types and dynamics . Social Compass , 58 , 498 – 510 . Google Scholar CrossRef Search ADS Tanaka K. ( 2010 ). Limitations for measuring religion in a different cultural context—the case of Japan . The Social Science Journal , 47 , 845 – 852 . Google Scholar CrossRef Search ADS PubMed Welch H. ( 1972 ). Buddhism under Mao . Cambridge, MA : Harvard University Press . Yang F. , Hu A. ( 2012 ). Mapping Folk Religion in China . Journal for the Scientific Study of Religion , 51 , 505 – 521 . Google Scholar CrossRef Search ADS Zhai J. ( 2010 ). Contrasting trends of religious markets in contemporary Mainland China and Taiwan . Journal of Church and State , 52 , 94s – 111 . Google Scholar CrossRef Search ADS APPENDIX Table A1 The Wordings of the Survey Items Conversion Did you ever go through a formal conversion ritual to become a Buddhist, such as the Buddhist conversion ceremony (guiyi)? (1 = yes; 0 = no) Existence Do you think Buddha and Bodhisattva exist? (1 = yes; 0 = no) Belief No matter whether you ever visited a church, temple, or other types of religious sites, do you have a belief in Buddhism? (1 = yes; 0 = no) Temple Which of the following activities did you participate in over the past year? Visiting a Buddhist temple to burn incense for Buddha and Bodhisattva (1 = yes; 0 = no) Worship Over the past 12 months, did you pay respect to Buddha and Bodhisattva? (1 = yes; 0 = no) Sutra Over the past 12 months, did you read/study Buddhist scriptures? (1 = yes; 0 = no) Conversion Did you ever go through a formal conversion ritual to become a Buddhist, such as the Buddhist conversion ceremony (guiyi)? (1 = yes; 0 = no) Existence Do you think Buddha and Bodhisattva exist? (1 = yes; 0 = no) Belief No matter whether you ever visited a church, temple, or other types of religious sites, do you have a belief in Buddhism? (1 = yes; 0 = no) Temple Which of the following activities did you participate in over the past year? Visiting a Buddhist temple to burn incense for Buddha and Bodhisattva (1 = yes; 0 = no) Worship Over the past 12 months, did you pay respect to Buddha and Bodhisattva? (1 = yes; 0 = no) Sutra Over the past 12 months, did you read/study Buddhist scriptures? (1 = yes; 0 = no) Table A1 The Wordings of the Survey Items Conversion Did you ever go through a formal conversion ritual to become a Buddhist, such as the Buddhist conversion ceremony (guiyi)? (1 = yes; 0 = no) Existence Do you think Buddha and Bodhisattva exist? (1 = yes; 0 = no) Belief No matter whether you ever visited a church, temple, or other types of religious sites, do you have a belief in Buddhism? (1 = yes; 0 = no) Temple Which of the following activities did you participate in over the past year? Visiting a Buddhist temple to burn incense for Buddha and Bodhisattva (1 = yes; 0 = no) Worship Over the past 12 months, did you pay respect to Buddha and Bodhisattva? (1 = yes; 0 = no) Sutra Over the past 12 months, did you read/study Buddhist scriptures? (1 = yes; 0 = no) Conversion Did you ever go through a formal conversion ritual to become a Buddhist, such as the Buddhist conversion ceremony (guiyi)? (1 = yes; 0 = no) Existence Do you think Buddha and Bodhisattva exist? (1 = yes; 0 = no) Belief No matter whether you ever visited a church, temple, or other types of religious sites, do you have a belief in Buddhism? (1 = yes; 0 = no) Temple Which of the following activities did you participate in over the past year? Visiting a Buddhist temple to burn incense for Buddha and Bodhisattva (1 = yes; 0 = no) Worship Over the past 12 months, did you pay respect to Buddha and Bodhisattva? (1 = yes; 0 = no) Sutra Over the past 12 months, did you read/study Buddhist scriptures? (1 = yes; 0 = no) Table A2 Comparison of Nested Confirmatory Factor Analytic Models With Latent Distinctions of the Three Bs for Evaluating Invariance by Demographic Variables for the Measures of Buddhism Involvement Model # Free parameters Chi-square value Chi-square DF (degree of freedom) Chi-square p-value CFI ΔCFI TLI RMSEA estimate RMSEA 90% CI lower bound RMSEA 90% CI higher bound RMSEA p-value Region of residence     1. Configural model 27 43.004 15 .0002 .998 .995 .023 .015 .031 1.000     2. Metric model 22 45.746 20 .0009 .998 .000 .997 .019 .012 .027 1.000     3. Scalar model 16 67.432 26 .0000 .996 −.002 .996 .021 .015 .028 1.000 Education     1. Configural model 27 45.348 15 .0001 .997 .995 .024 .016 .032 1.000     2. Metric model 22 49.749 20 .0003 .998 .001 .996 .020 .013 .028 1.000     3. Scalar model 16 59.833 26 .0002 .997 .000 .997 .019 .013 .026 1.000 Gender     1. Configural model 27 43.114 15 .0002 .997 .995 .023 .015 .021 1.000     2. Metric model 22 37.504 20 .0102 .998 .001 .998 .016 .008 .024 1.000     3. Scalar model 16 143.637 26 .0000 .989 −.008 .988 .036 .030 .042 1.000 Cohort     1. Configural model 27 43.004 15 .0000 .998 .995 .023 .015 .031 1.000     2. Metric model 22 45.746 20 .0009 .998 .000 .997 .019 .012 .027 1.000     3. Scalar model 16 67.432 26 .0000 .996 −.002 .996 .021 .015 .028 1.000 Model # Free parameters Chi-square value Chi-square DF (degree of freedom) Chi-square p-value CFI ΔCFI TLI RMSEA estimate RMSEA 90% CI lower bound RMSEA 90% CI higher bound RMSEA p-value Region of residence     1. Configural model 27 43.004 15 .0002 .998 .995 .023 .015 .031 1.000     2. Metric model 22 45.746 20 .0009 .998 .000 .997 .019 .012 .027 1.000     3. Scalar model 16 67.432 26 .0000 .996 −.002 .996 .021 .015 .028 1.000 Education     1. Configural model 27 45.348 15 .0001 .997 .995 .024 .016 .032 1.000     2. Metric model 22 49.749 20 .0003 .998 .001 .996 .020 .013 .028 1.000     3. Scalar model 16 59.833 26 .0002 .997 .000 .997 .019 .013 .026 1.000 Gender     1. Configural model 27 43.114 15 .0002 .997 .995 .023 .015 .021 1.000     2. Metric model 22 37.504 20 .0102 .998 .001 .998 .016 .008 .024 1.000     3. Scalar model 16 143.637 26 .0000 .989 −.008 .988 .036 .030 .042 1.000 Cohort     1. Configural model 27 43.004 15 .0000 .998 .995 .023 .015 .031 1.000     2. Metric model 22 45.746 20 .0009 .998 .000 .997 .019 .012 .027 1.000     3. Scalar model 16 67.432 26 .0000 .996 −.002 .996 .021 .015 .028 1.000 Note. CFI = comparative fit index; CI = confidence interval; RMSEA = root mean squared error of approximation TLI = Tucker–Lewis index. Source: Chinese Spiritual Life Survey 2007. Table A2 Comparison of Nested Confirmatory Factor Analytic Models With Latent Distinctions of the Three Bs for Evaluating Invariance by Demographic Variables for the Measures of Buddhism Involvement Model # Free parameters Chi-square value Chi-square DF (degree of freedom) Chi-square p-value CFI ΔCFI TLI RMSEA estimate RMSEA 90% CI lower bound RMSEA 90% CI higher bound RMSEA p-value Region of residence     1. Configural model 27 43.004 15 .0002 .998 .995 .023 .015 .031 1.000     2. Metric model 22 45.746 20 .0009 .998 .000 .997 .019 .012 .027 1.000     3. Scalar model 16 67.432 26 .0000 .996 −.002 .996 .021 .015 .028 1.000 Education     1. Configural model 27 45.348 15 .0001 .997 .995 .024 .016 .032 1.000     2. Metric model 22 49.749 20 .0003 .998 .001 .996 .020 .013 .028 1.000     3. Scalar model 16 59.833 26 .0002 .997 .000 .997 .019 .013 .026 1.000 Gender     1. Configural model 27 43.114 15 .0002 .997 .995 .023 .015 .021 1.000     2. Metric model 22 37.504 20 .0102 .998 .001 .998 .016 .008 .024 1.000     3. Scalar model 16 143.637 26 .0000 .989 −.008 .988 .036 .030 .042 1.000 Cohort     1. Configural model 27 43.004 15 .0000 .998 .995 .023 .015 .031 1.000     2. Metric model 22 45.746 20 .0009 .998 .000 .997 .019 .012 .027 1.000     3. Scalar model 16 67.432 26 .0000 .996 −.002 .996 .021 .015 .028 1.000 Model # Free parameters Chi-square value Chi-square DF (degree of freedom) Chi-square p-value CFI ΔCFI TLI RMSEA estimate RMSEA 90% CI lower bound RMSEA 90% CI higher bound RMSEA p-value Region of residence     1. Configural model 27 43.004 15 .0002 .998 .995 .023 .015 .031 1.000     2. Metric model 22 45.746 20 .0009 .998 .000 .997 .019 .012 .027 1.000     3. Scalar model 16 67.432 26 .0000 .996 −.002 .996 .021 .015 .028 1.000 Education     1. Configural model 27 45.348 15 .0001 .997 .995 .024 .016 .032 1.000     2. Metric model 22 49.749 20 .0003 .998 .001 .996 .020 .013 .028 1.000     3. Scalar model 16 59.833 26 .0002 .997 .000 .997 .019 .013 .026 1.000 Gender     1. Configural model 27 43.114 15 .0002 .997 .995 .023 .015 .021 1.000     2. Metric model 22 37.504 20 .0102 .998 .001 .998 .016 .008 .024 1.000     3. Scalar model 16 143.637 26 .0000 .989 −.008 .988 .036 .030 .042 1.000 Cohort     1. Configural model 27 43.004 15 .0000 .998 .995 .023 .015 .031 1.000     2. Metric model 22 45.746 20 .0009 .998 .000 .997 .019 .012 .027 1.000     3. Scalar model 16 67.432 26 .0000 .996 −.002 .996 .021 .015 .028 1.000 Note. CFI = comparative fit index; CI = confidence interval; RMSEA = root mean squared error of approximation TLI = Tucker–Lewis index. Source: Chinese Spiritual Life Survey 2007. Figure A1 View largeDownload slide Result of confirmatory factor analysis with latent distinctions of the three Bs. Note: All factor loadings are statistically significant at the 0.001 level. CI = confidence interval; bh. = behavior; bl. = believing; rel. = religiosity. Source: Chinese Spiritual Life Survey 2007. Figure A1 View largeDownload slide Result of confirmatory factor analysis with latent distinctions of the three Bs. Note: All factor loadings are statistically significant at the 0.001 level. CI = confidence interval; bh. = behavior; bl. = believing; rel. = religiosity. Source: Chinese Spiritual Life Survey 2007. © The Author 2017. Published by Oxford University Press on behalf of The World Association for Public Opinion Research. All rights reserved. http://www.deepdyve.com/assets/images/DeepDyve-Logo-lg.png International Journal of Public Opinion Research Oxford University Press

Evaluating Survey Items of Buddhism Religiosity in China

Loading next page...
 
/lp/ou_press/evaluating-survey-items-of-buddhism-religiosity-in-china-1chGpTVFL4
Publisher
Oxford University Press
Copyright
© The Author 2017. Published by Oxford University Press on behalf of The World Association for Public Opinion Research. All rights reserved.
ISSN
0954-2892
eISSN
1471-6909
D.O.I.
10.1093/ijpor/edx018
Publisher site
See Article on Publisher Site

Abstract

Over the past few decades, public opinion studies have paid increasing attention to the comparison of eastern and western religions, an endeavor that calls for a rigorous evaluation of empirical measures of religiosity for eastern religions. This study investigates the applicability to Chinese Buddhism of a measurement scheme based on believing, belonging, and behaving (the Three Bs). Confirmatory factor analyses of the Chinese Spiritual Life Survey suggest that the Three Bs are valid measures of Buddhist religiosity as they converge to the same latent construct. They are also reliable survey instruments, as illustrated by measurement invariance across subpopulations defined by differences in region of residence, gender, cohort, and education. The Three Bs measurement scheme has potential for empirical comparative research. Evaluating Survey Items of Buddhism Religiosity in China It has been noted that religion is strongly related to individual attitudes toward political parties, abortion, gay rights, and foreign policy, to name a few (Baumgartner, Francia, & Morris, 2008; Clements, 2014a; Eisenstein, 2006). With the rise in cross-national comparative studies, various religions, especially ones from the East, have begun to enter the research agenda of empirical researchers (Bender, Cadge, Levitt, & Smilde, 2012; Cadge, Levitt, & Smilde 2011; Gonzalez, 2011; Tanaka, 2010). This brings the following questions to the fore: How can researchers measure religious involvement in different religions using survey instruments? More importantly, is it possible to gauge religiosity in different religions with a common measurement scheme?1 To examine this question, this research note presents a case study that explores whether a measurement scheme originally developed for Christianity can be extended to the eastern religion of Buddhism in China. We chose Buddhism as our subject not only because of its well-documented worldwide presence and social influences (McMahan, 2011) but also because of its notable tendency to growth in traditionally Christian nations (Leamaster, 2012). In light of this, a meticulous evaluation of survey items pertaining to Buddhist religiosity serves the interest of comparative public opinion studies. We choose China as our research site because the two major challenges to the reliability and validity of the empirical measurement of religiosity in the east are loose organization and theological syncretism, and these are typical in Chinese society (Leamaster & Hu, 2014; Hu, 2016). By virtue of this case study, we find evidence for the validity and reliability of a common religiosity measurement scheme for Buddhism in China, which strengthens our confidence in the applicability of similar survey items in other countries. The Three Bs Measurement Scheme In this study, we draw on a religiosity measurement scheme that centers on the Three Bs: believing, behaving, and belonging. “Believing” describes a follower’s faith and confidence in a religion; “behaving” describes the practices and activities that are driven by one’s religious commitment; “belonging” refers to affiliation and identification with a particular religious tradition (Green, Guth, Smidt, & Kellstedt, 1996; Olson & Warber, 2008). The Three Bs scheme has mainly been applied to investigate Christianity, and it has been shown that each of the Three Bs reflects the same underlying latent construct of Christian religiosity (Clayton, 1971; Clayton & Gladden, 1974; de Jong, Faulkner, & Warland, 1976; Faulkner & de Jong, 1966). In addition, the Three Bs are recognized as the most fundamental and stable dimensions of Christian religiosity (Glock and Stark 1965).2 As a result, survey instruments concerning these three dimensions of religiosity have often been used together to investigate various empirical themes, such as parenting value (Sieben & Halman, 2014), public service participation (Freeman & Houston, 2010), opposition to abortion (Clements, 2014b), public opinion about same-sex marriage (Olson, Cadge, & Harrison, 2006), and voting behavior (Smith & Walker, 2012). Despite the wide-ranging utilization of the Three Bs measurement scheme in Christianity, it is not clear whether it can be also applied to Buddhism in the East. From the Durkheimian functionalist perspective on religion, we speculate that the religiosity of Buddhists, like that of Christians, can also be gauged by the Three Bs,3 as the essential elements of religion are ritual practices (behaving) and the ensuing beliefs and communal belonging (Marshall 2002). Before moving on to the empirical section, we first familiarize readers with the Chinese context, by highlighting potential challenges to the validity and reliability of survey items concerning Buddhist religiosity. The Case of Chinese Buddhism Challenges to Measurement Validity Measurement validity refers to whether the variations of empirical items are driven by the same theoretical construct (Gliner, Morgan, & Leech, 2009). A valid measurement scheme for religiosity should guarantee that all the empirical items reflect the same latent construct of religious propensity. This, however, might not be the case for the Three Bs because of some unique features of Chinese Buddhism. First, some Chinese citizens’ Buddhist identification may not be fully driven by religiosity. As suggested by Lu and Li (2011), one possible reason for some Chinese citizens’ identification with Buddhism is that they are not familiar with religions other than Buddhism, so, for them, belonging to Buddhism becomes a routine cultural identity. Second, the religious implication of Buddhist practices might be vague in the minds of some practitioners, as these practices are syncretic, insofar as being practiced by followers of multiple religions. This overlap of Buddhism with other Chinese religions (e.g., Daoism and folk religion) has been highlighted by scholars with the term “folk Buddhism” (Lin, 2008; Overmyer, 1976). Third, the utilitarian orientation of Chinese Buddhism means that believing in Buddhism in China is not necessarily an indicator of a person’s transcendental commitment and faith, but may rather be simply a means of satisfying one’s pragmatic pleas (Chau 2006; Nadeau, & Komjathy, 2012).4 In summary, extant studies on Chinese Buddhism indicate that the measurement items for behaving, believing, and belonging, when applied to Buddhism, may entail factors that are too diverse to converge on the same latent construct of religiosity, challenging measurement validity. Challenges to Measurement Reliability In addition to validity, studies on Chinese Buddhism imply some potential concerns about the reliability of the Three Bs. Reliability, defined as the consistency of measurement items, calls for measurement invariance across subgroups of respondents.5 In China, this property is confronted with challenges. For instance, females might be less active than males in joining formal Buddhist organizations (Cadge, 2004). Instead, they are more likely to be unorganized lay Buddhists (Fan, 2003). Therefore, Buddhist religiosity measures may reveal a gender difference, with women attaching more significance to believing and behaving rather than to belonging. Besides gender, measurement inconsistencies could emerge between educational groups. The better educated pay more attention to self-cultivation and the ethical meaning of Buddhism, such that they value believing more than behaving. In contrast, the less educated are more superstitious, so for them, Buddhist involvement gravitates toward the dimension of behaving (Hu & Leamaster, 2013). A third sociodemographic factor that might challenge the Three Bs’ reliability is cohort. Specifically, the generations who experienced the socialist regime in their formative stage might be reluctant to openly display their religious practices, but rather express their religious commitment through private belief. This is partly because of the fact that the state, during the socialist regime, suppressed people’s religious beliefs less efficiently than it did their overt religious activities (Lee 2007; Leung, 2005; Potter, 2003). As a result, an intergenerational inconsistency could emerge as the post-Reform generations have been subject to state repression to a much lesser extent. Finally, the urban–rural difference may be another source of measurement inconsistency. The commercialization of urban Buddhism encourages disciplinary laxity among clergies, resulting in a hollowing-out of the “plausibility” of urban Buddhism (Sun, 2011); therefore, urban residents’ belief and faith in Buddhism, relative to their rural counterparts, could be less firm (Oakes & Sutton, 2010). Altogether, prior literature indicates that the Three Bs measurement scheme, when applied to Chinese Buddhism, meets potential challenges in both survey validity and survey reliability that question its applicability to cross-national comparative public opinion research. In this research note, we rigorously respond to these concerns, and empirically investigate the quality of the Three Bs survey items. Data, Measures, and Analytical Strategy Sample This study analyzes data from the Chinese Spiritual Life Survey (CSLS), which was collected in 2007. The CSLS was the first representative nationwide survey with a focus on Chinese residents’ religious life. The total number of respondents of the CSLS was 7,021, with an age interval between 16 and 75 years. The survey was administered in 56 locales throughout mainland China. Within each locale, neighborhoods were sampled within administratively defined neighborhood committees (government-defined collections of neighborhoods), and households were randomly sampled within each neighborhood. A KISH grid procedure was used to randomly select one respondent for a face-to-face interview. The final sample had an American Association for Public Opinion Research (AAPOR) response rate of 28.1%.6 Although the CSLS was collected in 2007, it is to date the only survey that specifically focuses on the various aspects of Chinese citizens’ religious lives. Other than the CSLS, no other surveys (e.g., the Chinese General Social Survey conducted in 2010) provide all of the survey instruments of the Three Bs. Since being collected, the CSLS has been used by scholars to study a wide range of topics pertaining to Chinese religions (Liu and Mencken, 2010, Stark and Liu, 2011, Yang and Hu, 2012, and Zhai, 2010). Measures Both the Chinese and English wordings of the items are presented in Table A1. To shed light on comparative studies, we selected the most common measurement items for Buddhist religiosity that are applicable, as much as possible, to the various schools of Buddhism. Because of this, some survey items, which are more school-specific, were not used.7 A common practice in studies of Christianity is to measure belonging with denominational affiliation (e.g., Baptist or Methodist). However, no such nuanced denominational distinctions apply in Chinese Buddhism. Moreover, even if we distinguish between schools of Buddhism, a person simply considering himself/herself affiliated with one particular Buddhist tradition (e.g., the Mahayana) is insufficient in revealing his/her organizational attachment because of the syncretic nature of religious life and the vague meaning of religious affiliation in the Chinese context (Leamaster & Hu, 2014). Because of this, we argue that an item that can better gauge one’s “belonging” to Buddhism would be a stricter indicator of organizational involvement. Specifically, we measure Buddhist “belonging” using the question of whether a respondent converted to Buddhism through a formal guiyi ritual (conversion) (1 = yes; 0 = no).8 According to the theology of Buddhism, those who go through the guiyi ritual are considered to have taken refuge in the Three Jewels (Buddha, Dharma, and Sangha) (Kurtz, 2007; Welch, 1972). As noted by Hu and Leamaster (2013), the guiyi ritual implies an extensive and somewhat exclusivist commitment to Buddhism, so it should serve better than self-reported affiliation as a measure of one’s organizational belonging to Buddhism.9 The dimension of “believing” concerns people’s faith and confidence in Buddhism; this is measured by two questions (1 = yes; 0 = no). The first question directly asks whether one believes in Buddhism (belief) from a list of religions (Buddhism, Catholicism, Protestantism, Taoism, Islam, and others).10 Note that, in this study, we view this variable as an item relevant to the dimension of “believing” instead of “belonging” because the wording of this question gravitates more toward belief in (xin) Buddhism rather than organizational religious affiliation. As an individual’s attitudes toward the existence of spiritual figures in a religion represent a fundamental aspect of religious faith (Polkinghorne, 1998), our second measure for “believing” is a question about whether one believes that Buddha (fozu) and Bodhisattva (pusa) exist (existence) (1 = yes; 0 = no). It is worth mentioning that this question measures people’s Buddhist belief instead of Chinese folk religious belief because (1) the wording of the questionnaire made it clear that the subjects in question are fozu and pusa, two Buddhist figures; and (2) folk religious deities and spirits had already been listed elsewhere in the CSLS.11 Buddhist practices in China are diverse, such that it is almost impossible to exhaustively consider all of them. In this study, we chose some typical practices considered in the CSLS to measure “behaving” in Buddhism. Specifically, three questions were used, including whether participants burnt incense for Buddha and Bodhisattva in a Buddhist temple (temple), whether respondents paid respects to Buddha and Bodhisattva (worship), and whether they read or study Buddhist scriptures (sutra) (1 = yes; 0 = no). These questions represent the most common Buddhist practices in China, involving both private and public, as well as both organizational and nonorganizational activities (Birnbaum, 2003).12 The CSLS is, to date, the only large-scale survey that focuses exclusively on Chinese citizens’ religious life, but there is still room for further improvement. For example, the CSLS does not provide information about the more subtle variations between different traditions of Chinese Buddhism (e.g., the Vajrayana school of Buddhism) and other Buddhism-inspired practices. Hence, the items used here capture Buddhist involvement in the general sense. Also, as a study based on secondary data, we were not able to design our own items covering everything related to Buddhist involvement, so it is inevitable that the examined items are relatively simple in both form and content (e.g., several binary items). Finally, the self-presentation effect and social-expectation bias cannot be fully ruled out. For example, people who do not claim Buddhist “belonging,” “behaving,” or “believing” may still be involved in them in private.13 Notwithstanding these limitations, however, we believe that a formal evaluative study on the performance of the measurement instruments of the Three Bs for the case of Buddhism, one major religion in the East, is still meaningful, necessary, and timely in the light of increasing attention toward eastern religions in public opinion research. Analytical Strategy As stated earlier, empirical items are valid if they quantify the same underlying theoretical construct. To assess the validity of the Three Bs, we performed a confirmatory factor analysis (CFA) in which a single latent variable of Buddhist religiosity was configured for all of the observed items. In addition to this model, we also configured a CFA model in which multiple latent constructs were configured to, respectively, correspond to religious belonging, believing, and behaving.14 This more complicated model showed good model fit, but it was not significantly better than the one with a single latent variable.15 Because of this, we reported the CFA model in which all items were mapped to the single latent construct. Reliability was examined by considering the measurement invariance of the Three Bs items across sociodemographic groups, which was performed through multigroup confirmatory factor analysis (MGCFA). In MGCFA, there are three levels of measurement invariance (Meredith, 1993). The basic level is configural invariance, meaning that one model configuration can fit the data of different groups and that all item loadings are statistically significant. The second level is metric invariance. This level constrains factor loadings to be identical across groups. The third level is scalar invariance, where both factor loadings and intercepts are constrained as invariant across groups. Scalar invariance is stricter than metric invariance, which is, in turn, stricter than configural invariance. In the following research, we tested the reliability of the Three Bs measurement items for Chinese Buddhism across subpopulations defined by gender (male vs. female), education (below high school vs. high school and above), region of residence (urban vs. rural), and cohort (pre-1978 cohort vs. post-1978 cohort).16 In the following analyses, multiple indexes of model fit were used, including the Chi-square value, the root mean squared error of approximation (RMSEA), the comparative fit index (CFI), and the Tucker–Lewis index (TLI). RMSEA values <.05 indicate excellent fit (Hu and Bentler, 1999). The range of CFI and TLI is from 0 (poor fit) to 1 (perfect fit), and good model fit should return a value >.95.17 Results Descriptive Patterns Descriptive patterns pertaining to Chinese Buddhism can be found in Table 1. As shown, around 2% of the respondents of the CSLS had formally converted to Buddhism through the ritual of guiyi. In total, 7% believed that Buddha and Bodhisattva exist, and 17% reported belief in the religion of Buddhism. In terms of Buddhist practices, 11% of Chinese adults had burnt incense for Buddha and Bodhisattva in a Buddhist temple; comparatively, fewer individuals had paid respect to Buddha and Bodhisattva (6%) or read or studied Buddhist scriptures (1%). Table 1 Descriptive Statistics Variable Description % SD Conversion Formally convert to Buddhism through the guiyi ceremony 2 0.12 Existence Believing that Buddha and Bodhisattva exist 7 0.26 Belief Believing in Buddhism 17 0.37 Temple Burning incense for Buddha and Bodhisattva in a Buddhist temple 11 0.31 Worship Paying respect to Buddha and Bodhisattva 6 0.23 Sutra Reading/studying Buddhist scriptures 1 0.1 N 7,021 Variable Description % SD Conversion Formally convert to Buddhism through the guiyi ceremony 2 0.12 Existence Believing that Buddha and Bodhisattva exist 7 0.26 Belief Believing in Buddhism 17 0.37 Temple Burning incense for Buddha and Bodhisattva in a Buddhist temple 11 0.31 Worship Paying respect to Buddha and Bodhisattva 6 0.23 Sutra Reading/studying Buddhist scriptures 1 0.1 N 7,021 Source: Chinese Spiritual Life Survey 2007. Table 1 Descriptive Statistics Variable Description % SD Conversion Formally convert to Buddhism through the guiyi ceremony 2 0.12 Existence Believing that Buddha and Bodhisattva exist 7 0.26 Belief Believing in Buddhism 17 0.37 Temple Burning incense for Buddha and Bodhisattva in a Buddhist temple 11 0.31 Worship Paying respect to Buddha and Bodhisattva 6 0.23 Sutra Reading/studying Buddhist scriptures 1 0.1 N 7,021 Variable Description % SD Conversion Formally convert to Buddhism through the guiyi ceremony 2 0.12 Existence Believing that Buddha and Bodhisattva exist 7 0.26 Belief Believing in Buddhism 17 0.37 Temple Burning incense for Buddha and Bodhisattva in a Buddhist temple 11 0.31 Worship Paying respect to Buddha and Bodhisattva 6 0.23 Sutra Reading/studying Buddhist scriptures 1 0.1 N 7,021 Source: Chinese Spiritual Life Survey 2007. Evaluating Measurement Validity The results of the CFA model are presented in Figure 1. All of the six items are significantly associated with the same latent construct of Buddhist religiosity (p < .05). Among those items, specifically, the factor with the highest loading is belief (3.031), which is followed by worship (1.508), temple (1.457), existence (1.361), sutra (1.290), and conversion (1.181). Figure 1 View largeDownload slide Result of confirmatory factor analysis. Note: All factor loadings are statistically significant at the 0.001 level. CI = confidence interval; Rel. = religiosity. Source: Chinese Spiritual Life Survey 2007 Figure 1 View largeDownload slide Result of confirmatory factor analysis. Note: All factor loadings are statistically significant at the 0.001 level. CI = confidence interval; Rel. = religiosity. Source: Chinese Spiritual Life Survey 2007 To evaluate this model's goodness of fit, a series of indexes is reported. Specifically, the Chi-square value of this CFA model is 43.695. Given the 9 degrees of freedom, this value is significant at the .0001 level. This result supports the goodness of fit of this model. In addition to the Chi-square value, CFI (.997), TLI (.995), and RMSEA (.023, with 90% confidence interval to be .017–.031) all suggest excellent model fit. In summary, a CFA model with all Three Bs instruments directed at the same latent construct of Buddhist religiosity fits the CSLS data well, so the validity of the Three Bs is confirmed.18 Evaluating Measurement Reliability The results of the MGCFA can be found in Table 2. One way to check the reliability of the Three Bs is to investigate the model fit indexes across the three levels of measurement invariance. If the CFA continues to fit the data as we move from configural invariance to metric invariance and then to scalar invariance, we have evidence of measurement consistency across subpopulations. Table 2 Comparison of Nested Confirmatory Factor Analytic Models for Evaluating Invariance by Demographic Variables for the Measures of Buddhism Involvement Model # Free parameters Chi-square value Chi-square DF Chi-square p-value CFI ΔCFI TLI RMSEA estimate RMSEA 90% CI lower bound RMSEA 90% CI higher bound RMSEA p-value Region of residence     1. Configural model 24 61.604 18 .0000 .996 .994 .026 .019 .034 1.000     2. Metric model 18 59.461 24 .0001 .997 .001 .996 .021 .014 .027 1.000     3. Scalar model 12 81.111 30 .0000 .996 .000 .996 .022 .016 .028 1.000 Education     1. Configural model 24 58.513 18 .0000 .997 .994 .025 .018 .033 1.000     2. Metric model 18 57.036 24 .0002 .997 .000 .996 .020 .013 .026 1.000     3. Scalar model 12 67.128 30 .0001 .997 .000 .997 .019 .013 .025 1.000 Gender     1. Configural model 24 54.206 18 .0000 .997 .995 .024 .017 .031 1.000     2. Metric model 18 48.397 24 .0023 .998 .001 .997 .017 .010 .024 1.000     3. Scalar model 12 193.58 30 .0000 .985 −.013 .985 .039 .034 .045 .999 Cohort     1. Configural model 23 61.337 19 .0000 .996 .994 .025 .018 .032 1.000     2. Metric model 18 59.109 24 .0000 .997 .001 .996 .020 .014 .027 1.000     3. Scalar model 12 6.578 30 .0000 .997 .000 .997 .017 .011 .023 1.000 Model # Free parameters Chi-square value Chi-square DF Chi-square p-value CFI ΔCFI TLI RMSEA estimate RMSEA 90% CI lower bound RMSEA 90% CI higher bound RMSEA p-value Region of residence     1. Configural model 24 61.604 18 .0000 .996 .994 .026 .019 .034 1.000     2. Metric model 18 59.461 24 .0001 .997 .001 .996 .021 .014 .027 1.000     3. Scalar model 12 81.111 30 .0000 .996 .000 .996 .022 .016 .028 1.000 Education     1. Configural model 24 58.513 18 .0000 .997 .994 .025 .018 .033 1.000     2. Metric model 18 57.036 24 .0002 .997 .000 .996 .020 .013 .026 1.000     3. Scalar model 12 67.128 30 .0001 .997 .000 .997 .019 .013 .025 1.000 Gender     1. Configural model 24 54.206 18 .0000 .997 .995 .024 .017 .031 1.000     2. Metric model 18 48.397 24 .0023 .998 .001 .997 .017 .010 .024 1.000     3. Scalar model 12 193.58 30 .0000 .985 −.013 .985 .039 .034 .045 .999 Cohort     1. Configural model 23 61.337 19 .0000 .996 .994 .025 .018 .032 1.000     2. Metric model 18 59.109 24 .0000 .997 .001 .996 .020 .014 .027 1.000     3. Scalar model 12 6.578 30 .0000 .997 .000 .997 .017 .011 .023 1.000 Note. CFI = comparative fit index; CI = confidence interval; RMSEA = root mean squared error of approximation TLI = Tucker–Lewis index. Source: Chinese Spiritual Life Survey 2007. Table 2 Comparison of Nested Confirmatory Factor Analytic Models for Evaluating Invariance by Demographic Variables for the Measures of Buddhism Involvement Model # Free parameters Chi-square value Chi-square DF Chi-square p-value CFI ΔCFI TLI RMSEA estimate RMSEA 90% CI lower bound RMSEA 90% CI higher bound RMSEA p-value Region of residence     1. Configural model 24 61.604 18 .0000 .996 .994 .026 .019 .034 1.000     2. Metric model 18 59.461 24 .0001 .997 .001 .996 .021 .014 .027 1.000     3. Scalar model 12 81.111 30 .0000 .996 .000 .996 .022 .016 .028 1.000 Education     1. Configural model 24 58.513 18 .0000 .997 .994 .025 .018 .033 1.000     2. Metric model 18 57.036 24 .0002 .997 .000 .996 .020 .013 .026 1.000     3. Scalar model 12 67.128 30 .0001 .997 .000 .997 .019 .013 .025 1.000 Gender     1. Configural model 24 54.206 18 .0000 .997 .995 .024 .017 .031 1.000     2. Metric model 18 48.397 24 .0023 .998 .001 .997 .017 .010 .024 1.000     3. Scalar model 12 193.58 30 .0000 .985 −.013 .985 .039 .034 .045 .999 Cohort     1. Configural model 23 61.337 19 .0000 .996 .994 .025 .018 .032 1.000     2. Metric model 18 59.109 24 .0000 .997 .001 .996 .020 .014 .027 1.000     3. Scalar model 12 6.578 30 .0000 .997 .000 .997 .017 .011 .023 1.000 Model # Free parameters Chi-square value Chi-square DF Chi-square p-value CFI ΔCFI TLI RMSEA estimate RMSEA 90% CI lower bound RMSEA 90% CI higher bound RMSEA p-value Region of residence     1. Configural model 24 61.604 18 .0000 .996 .994 .026 .019 .034 1.000     2. Metric model 18 59.461 24 .0001 .997 .001 .996 .021 .014 .027 1.000     3. Scalar model 12 81.111 30 .0000 .996 .000 .996 .022 .016 .028 1.000 Education     1. Configural model 24 58.513 18 .0000 .997 .994 .025 .018 .033 1.000     2. Metric model 18 57.036 24 .0002 .997 .000 .996 .020 .013 .026 1.000     3. Scalar model 12 67.128 30 .0001 .997 .000 .997 .019 .013 .025 1.000 Gender     1. Configural model 24 54.206 18 .0000 .997 .995 .024 .017 .031 1.000     2. Metric model 18 48.397 24 .0023 .998 .001 .997 .017 .010 .024 1.000     3. Scalar model 12 193.58 30 .0000 .985 −.013 .985 .039 .034 .045 .999 Cohort     1. Configural model 23 61.337 19 .0000 .996 .994 .025 .018 .032 1.000     2. Metric model 18 59.109 24 .0000 .997 .001 .996 .020 .014 .027 1.000     3. Scalar model 12 6.578 30 .0000 .997 .000 .997 .017 .011 .023 1.000 Note. CFI = comparative fit index; CI = confidence interval; RMSEA = root mean squared error of approximation TLI = Tucker–Lewis index. Source: Chinese Spiritual Life Survey 2007. Based on the analytical results presented in Table 2, measurement equivalence is supported for variations in region of residence. RMSEA (<.05), CFI (>.95) and TLI (>.95) all suggest an excellent model fit at all three levels of measurement invariance. This finding means that people living in rural and urban areas in contemporary China perceive the relation between observed measurement items and latent Buddhist religiosity consistently. Similar results are obtained for group variations in gender, education, and cohort, as all three levels of measurement invariance are met according to the model fit indices (RMSEA <.05, CFI >.95, and TLI >.95). Taken together, this means that a consistent empirical link between the Three Bs items and latent Buddhist religiosity will be produced, whether the Three Bs measurement scheme is applied to females or males, people with different educational attainments, or people of different generations.19 Concluding Remarks Religion has been cited by public opinion researchers to account for people’s attitudes toward a variety of social issues. Over the past few decades, cross-national comparative studies have been increasing, directing public opinion researchers’ attention to various religions in different parts of the world, especially the East. However, both the validity and reliability of survey instruments for religiosity are threatened by the peculiar national sociocultural context, necessitating a formal evaluation of these measurement schemes. In this research note, we have presented a case study of Buddhism in China. Empirical results suggest that a measurement scheme based on religious belonging, behaving, and believing is valid and able to gauge the different aspects of a single latent construct of Buddhist religiosity. The reliability of this scheme is empirically confirmed by the measurement equivalence across regional, gender, generational, and educational subpopulations. That is to say, people of different sociodemographic backgrounds in China view the link between the observed items and latent Buddhist religiosity consistently. These findings suggest that, even in a country like China where people’s religious life is loosely organized and syncretic, the survey items of the Three Bs, which have mainly been applied to Christianity, demonstrate the same underlying indicator of religiosity and return reliable information across sociodemographic subgroups. In light of this, we expect the Three Bs measurement scheme to be applicable in other nations. For example, in the United States, followers of Buddhism have been well organized. Even those from the Zen tradition form “a dharma of place” to enhance collective belonging. In addition, Buddhist beliefs have been a source of solace for many people, and Buddhist ritualistic practices, such as the “empowerment” from a mentor, the way to show respect and veneration for the bodhi tree or animistic deities, have also been prevalent (Coleman, 2001; Storhoff & Whalen-Bridge, 2010; Seager, 1999). In this regard, embodiments of all Three Bs are abundant, which set the stage for the application of our examined measurement scheme beyond the East. In fact, we speculate that the survey validity and reliability of the Three Bs items would be even better than what we document here because Buddhist practices, beliefs, and affiliation in the United States, can be less syncretic than those in China. This study is encouraging in showing that the religiosity of Buddhists in an Eastern society is similar to that of Christians, in that it can be empirically gauged by the same three components of religiosity. Thus, future cross-national and cross-religion comparative studies can be conducted in a more confident and more subtle fashion. For example, the effects on public opinion of different religions (e.g., Christianity vs. Buddhism) can be elaborated upon in terms of not only the general interreligion variations but also the precise component variations of the Three Bs. Clearly, new insights can be gleaned from this type of research. Anning Hu is Professor of Sociology at Fudan University, Shanghai, China. His research interests include the Sociology of Religion, Culture, Trust, and Social Research Methodology. This study was supported by the School of Social Development and Public Policy of Fudan University, the shuguang scholar program (17SG08), and the chuangxin group project of Fudan University (IDH3458007). Footnotes 1The term “religiosity” denotes a latent and multidimensional construct that describes how religious a person is. Following the social scientific research paradigm, this construct should be empirically measured by a battery of items. 2Specifically, the ideological dimension captures people's belief and confidence in a religion. The intellectual aspect of religiosity refers to one’s knowledge of religious doctrines. The ritualistic dimension focuses on overt religious behaviors. Religious experience is defined as the feelings and emotions attached to one’s religion. Finally, the consequential dimension refers to a religion’s various effects on people's secular life. 3According to Durkheim (1912), different religions, notwithstanding their distinct theologies and literatures, are defined in terms of the common functions of fostering social integration and a sense of unity. 4This could be the case for Chinese Buddhism, especially in light of the purported magical efficacy (lingyan) of many popular Buddhist deities (Lopez, 1996). For example, Bodhisattva (guanyin) is one of the most popular Buddhist deities in Chinese culture. It is widely believed that guanyin grants most blessings that are requested. 5This is shown through the coefficient invariance in the following confirmatory factor analysis. Measurement invariance does not mean that different people have the same values for the observed items. 6This number is somewhat deflated because 18% of all households sampled were unoccupied/empty when the interviewer visited the address. To ensure data quality, a postsurvey team was formed to double check over 20% of the interviews via telephone. 7For example, in an earlier version of this research note, we included the item of “Prayer.” However, as pointed out by the editor, this measure can be called into question for the Vajrayana schools of Buddhism in China, where the focus is on meditation and experience. Although we have good reasons to suspect that the type of Buddhism in the survey areas of the CSLS is mostly Chinese Buddhism, we believe a general measurement scheme better serves the empirical comparative research. For the empirical analysis, excluding the item of “Prayer” improved the model fit of the confirmatory factor analysis. 8Guiyi refers to a ritual process through which a Buddhist adherent establishes a mentor–disciple relationship with a Buddhist master. Usually, before guiyi is performed, a follower has to learn Buddhism informally with the master for some time. After conversion through guiyi, a Buddhist often lives in a temple and becomes intensively involved in the collective ceremonies and classes there. These Buddhists must also maintain a strict monastic lifestyle. 9In this regard, Buddhist guiyi looks similar to Christian baptism, but entails the additional expectation of a mentor–disciple tie. This partly accounts for the relatively low percentage of formally converted Buddhists in the CSLS. 10One potential concern related to this measure is the culture of polytheism in China (Gries, Su, & Schak, 2012). Nevertheless, this concern should not be severe because most respondents only chose one option, although multiple options were allowed. 11This question was designed to emphasize the spiritual aspect of Chinese Buddhism. Respondents should not misunderstand Buddha to be Shakyamuni, a historical figure. 12A potential concern about these measures is that they might be driven by the general Chinese cultural tradition instead of Buddhist religiosity. An empirical test of this concern is one objective of this study. If this concern proves significant, the Three Bs measurement scheme will reveal weak validity. 13In this regard, a survey like the CSLS might underestimate the popularity of Buddhism as well as its various dimensions in contemporary China. 14The diagram of this model can be found in Figure A1. 15The values of CFI, TLI, and RMSEA for the two models are close. Moreover, the difference in the Chi-square value is 5.64 (degree of freedom =2), which is not statistically significant. 16We chose 1978 as the demarcation point because the comprehensive Reform was initiated in that year. 17To accommodate the binary nature of the observed instruments, we used the mean and variance-adjusted weighted least squares (WLSMV) estimator in Mplus (Muthén & Muthén, 2007). The WLSMV has the merit of being free from the normality assumption for observed items, so it is always viewed to be the best option for research using categorical items (Brown, 2006). 18In Figure A1, we also present the goodness of fit of the CFA model with multiple latent variables, which is also well fitted, but not necessarily better than the model reported here. 19For the sake of cross-validation, we also present the results of the MGCFA with multiple latent constructs in Table A2. Again, measurement invariance is affirmed. References Baumgartner J. C. , Francia P. L. , Morris J. S. ( 2008 ). A clash of civilizations? The influence of religion on public opinion of US foreign policy in the Middle East . Political Research Quarterly , 61 , 171 – 179 . Google Scholar CrossRef Search ADS Bender C. , Cadge W. , Levitt P. , Smilde D. ( 2012 ). Religion on the edge: De-centering and re-centering the sociology of religion . New York, NY : Oxford University Press . Google Scholar CrossRef Search ADS Birnbaum R. ( 2003 ). Buddhist China at the century's turn . The China Quarterly , 174 , 428 – 450 . Google Scholar CrossRef Search ADS Brown T. ( 2006 ). Confirmatory factor analysis for applied research . New York, NY : Guildford . Cadge W. ( 2004 ). Gendered religious organizations: The case of theravada buddhism in America . Gender and Society , 18 , 777 – 793 . Google Scholar CrossRef Search ADS Cadge W. , Levitt P. , Smilde D. ( 2011 ). De-centering and re-centering: rethinking concepts and methods in the sociological study of religion . Journal for the Scientific Study of Religion , 50 , 437 – 449 . Google Scholar CrossRef Search ADS Chau A. Y. ( 2006 ). Miraculous response: Doing folk religion in contemporary China . Stanford, CA : Stanford University Press . Clayton R. R. , Gladden J. ( 1974 ). The five dimensions of religiosity: Toward demythologizing a sacred artifact . Journal for the Scientific Study of Religion , 13 , 135 – 143 . Google Scholar CrossRef Search ADS Clayton R. R. ( 1971 ). 5-D or 1? Journal for the Scientific Study of Religion , 10 , 37 – 40 . Google Scholar CrossRef Search ADS Clements B. ( 2014a ). Religion and public opinion in Britain . New York, NY : Palgrave Macmillan . Clements B. ( 2014b ). Religion and the sources of public opposition to abortion in Britain: The role of ‘belonging’, ‘behaving’ and ‘believing.’ Sociology , 48 , 369 – 386 . Google Scholar CrossRef Search ADS Coleman J. W. ( 2001 ). The New Buddhism: The western transformation of an ancient tradition . New York, NY : Oxford University Press . De Jong G. F. , Faulkner J. , Warland R. ( 1976 ). Dimensions of religiosity reconsidered; evidence from a cross-cultural study . Social Forces , 54 , 866 – 889 . Google Scholar CrossRef Search ADS Durkheim E. ( 1912 ). The elementary forms of religious life . Thousand Oak, CA : Sage Publications . Eisenstein M. ( 2006 ). Rethinking the relationship between religion and political tolerance in the US . Political Behavior , 28 : 327 – 348 . Google Scholar CrossRef Search ADS Emerson M. O. , Mirola W. , Monahan S. ( 2010 ). Religion matters: What sociology teaches us about religion in our world? New York, NY : Pearson . Emmons R. A. , Paloutzian R. ( 2003 ). The psychology of religion . Annual Review of Psychology , 54 , 377 – 402 . Google Scholar CrossRef Search ADS PubMed Fan H. ( 1997 ). Footbinding, feminism and freedom: The liberation of women's bodies in modern China . New York, NY : Routledge . Fan L. ( 2003 ). The cult of the silkworm mother as a core of local community religion in a North China Village: Field study in Zhiwuying, Baoding, Hebei. In Overmyer D. (Ed.), Religion in China today (pp. 53 – 66 ). New York, NY : Cambridge University Press . Faulkner J. E. , de Jong G. ( 1966 ). Religiosity in 5-D: An empirical analysis . Social Forces , 45 , 246 – 254 . Google Scholar CrossRef Search ADS Freeman P. K. , Houston D. J. ( 2010 ). Belonging, believing, and behaving: The religious character of public servants . Administration and Society , 42 , 694 – 719 . Google Scholar CrossRef Search ADS Froese P. ( 2008 ). The plot to kill god: Findings from the soviet experiment in secularization . Berkeley, CA : University of California Press . Google Scholar CrossRef Search ADS Fukuyama Y. ( 1961 ). The major dimensions of church membership . Review of Religious Research , 2 , 154 – 161 . Google Scholar CrossRef Search ADS Gliner J. A. , Morgan G. , Leech N. ( 2009 ). Research design and analysis in applied settings: An integrated approach to design and analysis . New York, NY : Routledge . Glock C. Y. , Stark R. ( 1965 ). Christian beliefs and anti-semitism . New York, NY : Harper and Row . Gonzalez A. ( 2011 ). Measuring religiosity in a majority Muslim context: Gender, religious salience, and religious experience among Kuwaiti college students—a research note . Journal for the Scientific Study of Religion , 50 , 339 – 350 . Google Scholar CrossRef Search ADS Green J. , Guth J. , Smidt C. , Kellstedt L. , ( 1996 ). Religion and the culture wars: Dispatches from the front . Lanham, MD : Rowman & Littlefield Publishers . Gries P. , Su J. , Schak D. ( 2012 ). Toward the scientific study of polytheism: Beyond forced‐choice measures of religious belief . Journal for the Scientific Study of Religion , 51 , 623 – 637 . Google Scholar CrossRef Search ADS Hu A. , Leamaster R. ( 2013 ). Longitudinal trends of religious. Groups in deregulated Taiwan: 1990-2009 . Sociological Quarterly , 54 , 254 – 277 . Google Scholar CrossRef Search ADS Hu A. ( 2016 ). Ancestor worship in contemporary China: An empirical investigation . China Review: An Interdisciplinary Journal on Greater China , 16 , 169 – 186 . Hu L. , Bentler P. ( 1999 ). Cutoff criteria for fit indexes in covariance structure analysis: Conventional criteria versus new alternatives . Structural Equation Modeling , 6 , 1 – 55 . Google Scholar CrossRef Search ADS Kurtz L. ( 2007 ). Gods in the global village: The world's religions in sociological perspective . Thousand Oaks, CA : Sage Publications . Leamaster R. , Hu A. ( 2014 ). Popular Buddhists: Predicting popular religious belief and practice in contemporary China . Sociology of Religion , 75 , 234 – 259 . Google Scholar CrossRef Search ADS Leamaster R. J. ( 2012 ). A research note on english‐speaking Buddhists in the United States . Journal for the Scientific Study of Religion , 51 , 143 – 155 . Google Scholar CrossRef Search ADS Lee T. ( 2007 ). Christianity in contemporary China: An update . Journal of Church and State , 49 , 277 – 304 . Google Scholar CrossRef Search ADS Leung B. ( 2005 ). China’s religious freedom policy: The art of managing religious activity . The China Quarterly , 184 , 894 – 913 . Google Scholar CrossRef Search ADS Lin M. ( 2008 ). Mazu Xinyang yu Hanren Shehui [Belief of Mazu and Han Chinese Society in Taiwan] . Ha-er-bing, China : Helongjiang Renmin Chubanshe . Liu E. , Mencken C. ( 2010 ). Fatalistic voluntarism and life happiness in post-socialist China . Sociological Spectrum , 30 , 270 – 288 . Google Scholar CrossRef Search ADS Lopez D. ( 1996 ). Religions of China in practice . Princeton, NJ : Princeton University Press . Lu Y. , Li D. ( 2011 ). The pattern, contemporary situation, and trends of Taiwan Religions. In Ji Z. , Qiu Y. (Eds.) The annual report on religions in China (pp. 225 – 252 ). Beijing, China : Chinese Social Sciences Press . Marshall D. ( 2002 ). Behavior, belonging, and belief: A theory of ritual practice . Sociological Theory , 20 , 360 – 380 . Google Scholar CrossRef Search ADS McMahan D. L. ( 2011 ). Buddhism in the Modern World . New York, NY : Routledge . Meredith W. ( 1993 ). Measurement invariance, factor analysis, and factorial invariance . Psychometrika , 58 , 525 – 543 . Google Scholar CrossRef Search ADS Muthén L. , Muthén B. ( 2007 ). Mplus user’s guide 4 . Los Angeles, CA : Author . Nadeau R. , Komjathy L. ( 2012 ). The wiley-blackwell companion to Chinese religions . New York, NY : Wiley-Blackwell . Google Scholar CrossRef Search ADS Oakes T. , Sutton D. ( 2010 ). Faiths on display: Religion, tourism, and the Chinese state . Lanham, MD : Rowman & Littlefield Publishers . Olson L. , Warber A. ( 2008 ). Belonging, behaving, and believing: Assessing the role of religion on presidential approval . Political Research Quarterly , 61 , 192 – 204 . Google Scholar CrossRef Search ADS Olson L. R. , Cadge W. , Harrison J. T. ( 2006 ). Religion and public opinion about same‐sex marriage . Social Science Quarterly , 87 , 340 – 360 . Google Scholar CrossRef Search ADS Overmyer D. ( 1976 ). Folk Buddhist religion: Dissenting sects in late traditional China . Cambridge, MA : Harvard University Press . Google Scholar CrossRef Search ADS Polkinghorne J. ( 1998 ). Belief in god in an age of science . New Haven, CT : Yale University Press . Potter P. ( 2003 ). Belief in control: Regulation of religion in China . The China Quarterly , 174 , 317 – 337 . Seager R. H. ( 1999 ). Buddhism in America . New York, NY : Columbia University Press . Sieben I. , Halman L. ( 2014 ). Religion and parental values in a secularized country: Evidence from the Netherlands . Social Compass , 61 , 121 – 140 . Google Scholar CrossRef Search ADS Smith L. , Walker L. ( 2012 ). Belonging, believing, and group behavior: Religiosity and voting in American Presidential Elections . Political Research Quarterly , 66 , 399 – 413 . Google Scholar CrossRef Search ADS Stark R. , Glock C. ( 1968 ). Patterns of religious commitment . Berkeley, CA : University of California Press . Stark R. , Liu E. ( 2011 ). The religious awakening in China . Review of Religious Research , 52 , 282 – 289 . Storhoff G. , Whalen-Bridge J. ( 2010 ). American Buddhism as a Way of Life . New York, NY : SUNY Press . Sun Y. ( 2011 ). The Chinese Buddhist ecology in post-mao China: Contours, types and dynamics . Social Compass , 58 , 498 – 510 . Google Scholar CrossRef Search ADS Tanaka K. ( 2010 ). Limitations for measuring religion in a different cultural context—the case of Japan . The Social Science Journal , 47 , 845 – 852 . Google Scholar CrossRef Search ADS PubMed Welch H. ( 1972 ). Buddhism under Mao . Cambridge, MA : Harvard University Press . Yang F. , Hu A. ( 2012 ). Mapping Folk Religion in China . Journal for the Scientific Study of Religion , 51 , 505 – 521 . Google Scholar CrossRef Search ADS Zhai J. ( 2010 ). Contrasting trends of religious markets in contemporary Mainland China and Taiwan . Journal of Church and State , 52 , 94s – 111 . Google Scholar CrossRef Search ADS APPENDIX Table A1 The Wordings of the Survey Items Conversion Did you ever go through a formal conversion ritual to become a Buddhist, such as the Buddhist conversion ceremony (guiyi)? (1 = yes; 0 = no) Existence Do you think Buddha and Bodhisattva exist? (1 = yes; 0 = no) Belief No matter whether you ever visited a church, temple, or other types of religious sites, do you have a belief in Buddhism? (1 = yes; 0 = no) Temple Which of the following activities did you participate in over the past year? Visiting a Buddhist temple to burn incense for Buddha and Bodhisattva (1 = yes; 0 = no) Worship Over the past 12 months, did you pay respect to Buddha and Bodhisattva? (1 = yes; 0 = no) Sutra Over the past 12 months, did you read/study Buddhist scriptures? (1 = yes; 0 = no) Conversion Did you ever go through a formal conversion ritual to become a Buddhist, such as the Buddhist conversion ceremony (guiyi)? (1 = yes; 0 = no) Existence Do you think Buddha and Bodhisattva exist? (1 = yes; 0 = no) Belief No matter whether you ever visited a church, temple, or other types of religious sites, do you have a belief in Buddhism? (1 = yes; 0 = no) Temple Which of the following activities did you participate in over the past year? Visiting a Buddhist temple to burn incense for Buddha and Bodhisattva (1 = yes; 0 = no) Worship Over the past 12 months, did you pay respect to Buddha and Bodhisattva? (1 = yes; 0 = no) Sutra Over the past 12 months, did you read/study Buddhist scriptures? (1 = yes; 0 = no) Table A1 The Wordings of the Survey Items Conversion Did you ever go through a formal conversion ritual to become a Buddhist, such as the Buddhist conversion ceremony (guiyi)? (1 = yes; 0 = no) Existence Do you think Buddha and Bodhisattva exist? (1 = yes; 0 = no) Belief No matter whether you ever visited a church, temple, or other types of religious sites, do you have a belief in Buddhism? (1 = yes; 0 = no) Temple Which of the following activities did you participate in over the past year? Visiting a Buddhist temple to burn incense for Buddha and Bodhisattva (1 = yes; 0 = no) Worship Over the past 12 months, did you pay respect to Buddha and Bodhisattva? (1 = yes; 0 = no) Sutra Over the past 12 months, did you read/study Buddhist scriptures? (1 = yes; 0 = no) Conversion Did you ever go through a formal conversion ritual to become a Buddhist, such as the Buddhist conversion ceremony (guiyi)? (1 = yes; 0 = no) Existence Do you think Buddha and Bodhisattva exist? (1 = yes; 0 = no) Belief No matter whether you ever visited a church, temple, or other types of religious sites, do you have a belief in Buddhism? (1 = yes; 0 = no) Temple Which of the following activities did you participate in over the past year? Visiting a Buddhist temple to burn incense for Buddha and Bodhisattva (1 = yes; 0 = no) Worship Over the past 12 months, did you pay respect to Buddha and Bodhisattva? (1 = yes; 0 = no) Sutra Over the past 12 months, did you read/study Buddhist scriptures? (1 = yes; 0 = no) Table A2 Comparison of Nested Confirmatory Factor Analytic Models With Latent Distinctions of the Three Bs for Evaluating Invariance by Demographic Variables for the Measures of Buddhism Involvement Model # Free parameters Chi-square value Chi-square DF (degree of freedom) Chi-square p-value CFI ΔCFI TLI RMSEA estimate RMSEA 90% CI lower bound RMSEA 90% CI higher bound RMSEA p-value Region of residence     1. Configural model 27 43.004 15 .0002 .998 .995 .023 .015 .031 1.000     2. Metric model 22 45.746 20 .0009 .998 .000 .997 .019 .012 .027 1.000     3. Scalar model 16 67.432 26 .0000 .996 −.002 .996 .021 .015 .028 1.000 Education     1. Configural model 27 45.348 15 .0001 .997 .995 .024 .016 .032 1.000     2. Metric model 22 49.749 20 .0003 .998 .001 .996 .020 .013 .028 1.000     3. Scalar model 16 59.833 26 .0002 .997 .000 .997 .019 .013 .026 1.000 Gender     1. Configural model 27 43.114 15 .0002 .997 .995 .023 .015 .021 1.000     2. Metric model 22 37.504 20 .0102 .998 .001 .998 .016 .008 .024 1.000     3. Scalar model 16 143.637 26 .0000 .989 −.008 .988 .036 .030 .042 1.000 Cohort     1. Configural model 27 43.004 15 .0000 .998 .995 .023 .015 .031 1.000     2. Metric model 22 45.746 20 .0009 .998 .000 .997 .019 .012 .027 1.000     3. Scalar model 16 67.432 26 .0000 .996 −.002 .996 .021 .015 .028 1.000 Model # Free parameters Chi-square value Chi-square DF (degree of freedom) Chi-square p-value CFI ΔCFI TLI RMSEA estimate RMSEA 90% CI lower bound RMSEA 90% CI higher bound RMSEA p-value Region of residence     1. Configural model 27 43.004 15 .0002 .998 .995 .023 .015 .031 1.000     2. Metric model 22 45.746 20 .0009 .998 .000 .997 .019 .012 .027 1.000     3. Scalar model 16 67.432 26 .0000 .996 −.002 .996 .021 .015 .028 1.000 Education     1. Configural model 27 45.348 15 .0001 .997 .995 .024 .016 .032 1.000     2. Metric model 22 49.749 20 .0003 .998 .001 .996 .020 .013 .028 1.000     3. Scalar model 16 59.833 26 .0002 .997 .000 .997 .019 .013 .026 1.000 Gender     1. Configural model 27 43.114 15 .0002 .997 .995 .023 .015 .021 1.000     2. Metric model 22 37.504 20 .0102 .998 .001 .998 .016 .008 .024 1.000     3. Scalar model 16 143.637 26 .0000 .989 −.008 .988 .036 .030 .042 1.000 Cohort     1. Configural model 27 43.004 15 .0000 .998 .995 .023 .015 .031 1.000     2. Metric model 22 45.746 20 .0009 .998 .000 .997 .019 .012 .027 1.000     3. Scalar model 16 67.432 26 .0000 .996 −.002 .996 .021 .015 .028 1.000 Note. CFI = comparative fit index; CI = confidence interval; RMSEA = root mean squared error of approximation TLI = Tucker–Lewis index. Source: Chinese Spiritual Life Survey 2007. Table A2 Comparison of Nested Confirmatory Factor Analytic Models With Latent Distinctions of the Three Bs for Evaluating Invariance by Demographic Variables for the Measures of Buddhism Involvement Model # Free parameters Chi-square value Chi-square DF (degree of freedom) Chi-square p-value CFI ΔCFI TLI RMSEA estimate RMSEA 90% CI lower bound RMSEA 90% CI higher bound RMSEA p-value Region of residence     1. Configural model 27 43.004 15 .0002 .998 .995 .023 .015 .031 1.000     2. Metric model 22 45.746 20 .0009 .998 .000 .997 .019 .012 .027 1.000     3. Scalar model 16 67.432 26 .0000 .996 −.002 .996 .021 .015 .028 1.000 Education     1. Configural model 27 45.348 15 .0001 .997 .995 .024 .016 .032 1.000     2. Metric model 22 49.749 20 .0003 .998 .001 .996 .020 .013 .028 1.000     3. Scalar model 16 59.833 26 .0002 .997 .000 .997 .019 .013 .026 1.000 Gender     1. Configural model 27 43.114 15 .0002 .997 .995 .023 .015 .021 1.000     2. Metric model 22 37.504 20 .0102 .998 .001 .998 .016 .008 .024 1.000     3. Scalar model 16 143.637 26 .0000 .989 −.008 .988 .036 .030 .042 1.000 Cohort     1. Configural model 27 43.004 15 .0000 .998 .995 .023 .015 .031 1.000     2. Metric model 22 45.746 20 .0009 .998 .000 .997 .019 .012 .027 1.000     3. Scalar model 16 67.432 26 .0000 .996 −.002 .996 .021 .015 .028 1.000 Model # Free parameters Chi-square value Chi-square DF (degree of freedom) Chi-square p-value CFI ΔCFI TLI RMSEA estimate RMSEA 90% CI lower bound RMSEA 90% CI higher bound RMSEA p-value Region of residence     1. Configural model 27 43.004 15 .0002 .998 .995 .023 .015 .031 1.000     2. Metric model 22 45.746 20 .0009 .998 .000 .997 .019 .012 .027 1.000     3. Scalar model 16 67.432 26 .0000 .996 −.002 .996 .021 .015 .028 1.000 Education     1. Configural model 27 45.348 15 .0001 .997 .995 .024 .016 .032 1.000     2. Metric model 22 49.749 20 .0003 .998 .001 .996 .020 .013 .028 1.000     3. Scalar model 16 59.833 26 .0002 .997 .000 .997 .019 .013 .026 1.000 Gender     1. Configural model 27 43.114 15 .0002 .997 .995 .023 .015 .021 1.000     2. Metric model 22 37.504 20 .0102 .998 .001 .998 .016 .008 .024 1.000     3. Scalar model 16 143.637 26 .0000 .989 −.008 .988 .036 .030 .042 1.000 Cohort     1. Configural model 27 43.004 15 .0000 .998 .995 .023 .015 .031 1.000     2. Metric model 22 45.746 20 .0009 .998 .000 .997 .019 .012 .027 1.000     3. Scalar model 16 67.432 26 .0000 .996 −.002 .996 .021 .015 .028 1.000 Note. CFI = comparative fit index; CI = confidence interval; RMSEA = root mean squared error of approximation TLI = Tucker–Lewis index. Source: Chinese Spiritual Life Survey 2007. Figure A1 View largeDownload slide Result of confirmatory factor analysis with latent distinctions of the three Bs. Note: All factor loadings are statistically significant at the 0.001 level. CI = confidence interval; bh. = behavior; bl. = believing; rel. = religiosity. Source: Chinese Spiritual Life Survey 2007. Figure A1 View largeDownload slide Result of confirmatory factor analysis with latent distinctions of the three Bs. Note: All factor loadings are statistically significant at the 0.001 level. CI = confidence interval; bh. = behavior; bl. = believing; rel. = religiosity. Source: Chinese Spiritual Life Survey 2007. © The Author 2017. Published by Oxford University Press on behalf of The World Association for Public Opinion Research. All rights reserved.

Journal

International Journal of Public Opinion ResearchOxford University Press

Published: Dec 13, 2017

There are no references for this article.

You’re reading a free preview. Subscribe to read the entire article.


DeepDyve is your
personal research library

It’s your single place to instantly
discover and read the research
that matters to you.

Enjoy affordable access to
over 18 million articles from more than
15,000 peer-reviewed journals.

All for just $49/month

Explore the DeepDyve Library

Search

Query the DeepDyve database, plus search all of PubMed and Google Scholar seamlessly

Organize

Save any article or search result from DeepDyve, PubMed, and Google Scholar... all in one place.

Access

Get unlimited, online access to over 18 million full-text articles from more than 15,000 scientific journals.

Your journals are on DeepDyve

Read from thousands of the leading scholarly journals from SpringerNature, Elsevier, Wiley-Blackwell, Oxford University Press and more.

All the latest content is available, no embargo periods.

See the journals in your area

DeepDyve

Freelancer

DeepDyve

Pro

Price

FREE

$49/month
$360/year

Save searches from
Google Scholar,
PubMed

Create lists to
organize your research

Export lists, citations

Read DeepDyve articles

Abstract access only

Unlimited access to over
18 million full-text articles

Print

20 pages / month

PDF Discount

20% off