Studying Religiosity and Spirituality: A Review of Macro, Micro, and Meso-Level Approaches

This paper seeks to advance the global study of religiosity and spirituality by conducting a meta-analysis of major approaches in the field. While the field, and thus the collected publications, are dominated by Western approaches, particular attention is paid in this analysis to publications from geographies that are not from the United States or Western Europe, especially these world regions: Africa, Asia, and Latin America. Similarly, while the study of religiosity is considerably centered around Christianity, this analysis extends beyond Christianity, to the extent possible in extant studies, to include publications investigating other world religious traditions, such as African spirituality, African witchcraft, Afro-Caribbean religious traditions, Buddhism, Confucianism, folk religions, Hinduism, Islam, Judaism, Mormonism, Neo-paganism, New Religious Movements (NRMs), Shamanism, Sikhism, Spiritism, Taoism, and spirituality generally. A total of 530 publications were reviewed, and the studies are categorized by unit of analysis into: Macro, micro, and meso-level. Measurement constructs include religious demography, culture, belonging, behaving, believing, bonding, religious salience, spiritual identities, religious networks, occupations, congregations, denominations, and faith-based organizations. Non-Western sources and approaches are analyzed toward furthering future research in under-studied world regions. Implications are drawn for the field, such as the need to geo-code publications at the country level.


Introduction
The aim of this paper is to advance the global study of religiosity and spirituality. To do so, the paper targets four major barriers to this goal. First, the majority of studies on religiosity and spirituality are based in Northern America or Western Europe (e.g., Bender 2013). This is a major impediment because Western assumptions, approaches, and biases are embedded within the theories, study designs, methods of data collection, and ways of operationalizing key constructs. Of particular concern is that the language that is often employed, even when translated, embeds particular approaches to the study of religion that are not sensitive to, or applicable within, under-studied world regions, and thus ignores the important role of local context (e.g., . For example, in the United States, one of the most heavily used measures for religiosity is individual frequency of attendance at religious services (e.g., Cadge et al. 2011;Smilde 2010). However, this focuses on a formalized, organizational version of religiosity that is not necessarily reflective of the informal spirituality practiced in many parts of the world. A methodological artifact of this Western and formal measure is that the true presence and meaning of religiosity and spirituality is undercounted and misunderstood in under-studied world regions, such as Africa and Asia. Thus, a simple count of adherents based on affiliation may severely underrepresent the true presence of faith and spirituality within cultures that are less congregational.
Second, a related issue is the heavy emphasis on Christian religions, which are dominant within the most well-studied world regions. Since most research on religiosity and spirituality occurs within Northern America and Western Europe, it follows that most of the measures have been developed to focus on the study of Christian denominations within those regions. Researchers of and within Western contexts often state that the developed measures can be generalized to non-Abrahamic faiths, but the empirical tests of this are often based within Christian-dominant countries. This approach to testing generalizability has major limitations because the sample size of non-Christian religions in Western contexts is typically small. Even when over-sampling approaches are employed, the minority religions in Western contexts may not be representative of non-Christian faiths based in other world regions, where Christianity is not as dominant.
Third, Western and Christian biases aside, another major issue impeding advancement in the broader study of religiosity and spirituality is the lack of a relatively cohesive field of study. The multi-disciplinarity that characterizes the field brings diverse sets of approaches that contribute rich and plural forms of knowledge construction processes. Yet, the depth and breadth of the accumulated data are rarely summarized and categorized. The field lacks sufficient meta-analyses that adequately reference the array of existing approaches. Instead, the literature reviews of most journal articles attend only to the set of studies that are most similar to the employed method. Even when researchers cite different constructs, or alternative methods, the level of analysis is typically the same. For example, a study of individual religiosity is most likely to survey other micro-level studies, and a cross-national comparison most likely to focus on macro-level studies. Rarely are current publications summarizing insights garnered across multiple levels of study. This is problematic because the depth of insights garnered at one level are not shared at another level. For example, micro-level studies have produced numerous ways of investigating individual religiosity and spirituality with greater depth than attention to denominational affiliation alone, but many macro-level studies continue to compare nations based on this sole measure of religious affiliation.
Fourth, archaic divisions over methodological paradigms inhibited data accumulation. Rather than valuing pluralistic methods that amass data from multiple sources of input, quantitative studies were divorced from qualitative and interpretive methods. Contemporary studies more often overcome this division by employing mixed methods to data collection within a single study. While this within-study approach advances the field in important ways, more attention needs to be paid to between-study method pluralism, by extracting learnings garnered from qualitative results in the design of quantitative studies, and in the reverse to relying on qualitative studies to interpret and contextualize quantitative results. This is particularly necessary when accumulating knowledge from the limited number of studies existing within under-researched regions of the world, many of which Religions 2020, 11, 437 4 of 74 are qualitative and interpretive. Separating this highly meaningful and contextualized information from the design of quantitative studies facilitates the ongoing Western and Christian biases within typical survey and experimental designs. Rectifying this issue requires attending to multiple methods and many sources of data, prioritizing information developed within and from the under-studied regions. Otherwise, the field risks perpetuating a de facto form of colonialism that continues to export Western measures to countries that may, for important reasons, not fit a standardized mold of pre-tested survey indicators.
To respond to these four barriers, this paper provides a review of major themes in the global study of religiosity and spirituality. The primary goal is to expand the geographic scope of the field to prioritize knowledge created by and gathered within under-studied world regions, especially Africa, Asia, and Latin America. This is accomplished by first summarizing the major approaches within each level of analysis: Macro, micro, and meso. For example, the macro-level section reviews studies of global religious demography that count adherents of major religious traditions by nation or continent, as well as other approaches that focus on nations as the key unit of analysis. Alternatively, the micro-level section summarizes approaches to studying individual religiosity and spirituality, paying particular attention to the "four B's": Belonging, behaving, believing, and bonding (e.g., Saroglou 2011), in conjunction with reviews of studies focused on spiritual identities. Third, the meso-level section reviews intersectional approaches that contextualize individuals within organizational units, such as religious congregations or faith-based organizations, which are shaped within larger macro-level social institutions, such as nations and denominations, the latter of which often span national boundaries. Accumulating the results across multiple levels of analysis reveals several areas of growth for future studies. For example, macro-level research could expand data collection beyond a heavy emphasis on affiliation. Meso-level reserch could assume the multi-affiliations found at the micro-level and develop more porous and fluid ways of understanding organizational boundaries.
More generally, summarizing the state of the field across each of the major levels of analysis provides an avenue for advancing future studies that attend to under-research world regions. Rather than risking replicating the same Christian biases in new geographies, scholarship developed within non-Western contexts needs to be prioritized and supported. Yet, scholars in under-studied world regions need not reinvent the same pitfalls, inadequacies, and misinterpretations that burdened prior decades of studies. Important information can be extracted from the myriad of existing studies toward designing non-Western approaches that attend to important dimensions of religiosity and spirituality across each level of analysis, and without historical divides between methods. The best first approach in a country with limited data on religiosity and spirituality is reflective and interpretive qualitative investigations. Only upon this strong foundation can a valid quantitative approach be scaled. Once quantitative measures are developed from within non-Western contexts, comparisons across nations and cultures can advance without the current risks for Western biases embedded in available metrics.

Method
To prioritize information published in a wide variety of sources, and using a range of methodological tools, this meta-analysis is based on a purposeful search (e.g., Suri 2011;Pawson 2006;Luciano 2011), conducted within expansive publication databases, such as EBSCO and Google Scholar. Due to major concerns over the likelihood of particular publication outlets or specified sources yielding the same Western and Christian centrism described above, this paper intentionally did not proscribe inclusion and exclusion parameters based on publication source. Rather, the researchers prioritized their collective domain knowledge to search sources from a wide array of inputs and then described the returned results. In this respect, the methods can be compared to a qualitative study based on inductive logic: The goal is to lessen potential researcher biases in pre-conceived ideas regarding what is counted as valid. Thus, prior to summarizing the results, the next section first provides descriptive statistics on the publication sources, scope, and contours.

Results
This section presents the results of the meta-analysis. Studies are categorized based on level of analysis: Macro, micro, and meso. The levels are visualized in Figure 3 and further described below.

Levels of Analysis
In contemporary scholarship, it is fairly common for scholars to categorize their unit of analysis as focusing on the micro, macro, or meso levels (e.g., Barman 2017;Corcoran et al. 2012;Hirschle 2013;Hustinx et al. 2010;Kregting et al. 2018;O'Brien and Noy 2018;Schafer and Upenieks 2016). As a brief description of these levels, micro-level approaches typically focus on individuals; meso-level approaches typically focus on organizations, groups, and networks; and macro-level approaches typically focus on large-scale social institutions, such as the government, economic, or legal system, often bounded at the nation-state level such that countries are the unit of analysis. A major reason for attending to these levels in contemporary scholarship is the increasing development of interdisciplinary approaches. Focusing on one versus another of these levels alone is characteristic of particular disciplines. For example, psychological and biopsychosocial studies regularly focus on the

Levels of Analysis
In contemporary scholarship, it is fairly common for scholars to categorize their unit of analysis as focusing on the micro, macro, or meso levels (e.g., Barman 2017;Corcoran et al. 2012;Hirschle 2013;Hustinx et al. 2010;Kregting et al. 2018;O'Brien and Noy 2018;Schafer and Upenieks 2016). As a brief description of these levels, micro-level approaches typically focus on individuals; meso-level approaches typically focus on organizations, groups, and networks; and macro-level approaches typically focus on large-scale social institutions, such as the government, economic, or legal system, often bounded at the nation-state level such that countries are the unit of analysis. A major reason for attending to these levels in contemporary scholarship is the increasing development of interdisciplinary approaches. Focusing on one versus another of these levels alone is characteristic of particular disciplines. For example, psychological and biopsychosocial studies regularly focus on the micro level, such as individuals. Alternatively, political science and economics regularly focus on the macro level, such as nation-states and gross domestic product (GDP). Between these, disciplines such as nonprofit studies, which itself is fairly interdisciplinary, focus on the meso level, such as nonprofit organizations. Other disciplines, such as sociology and religious studies, which also have interdisciplinary roots, span multiple levels, such as by studying familial, educational, and religious communities. Interdisciplinary scholarship cumulatively includes all three levels ( Figure 3).  Trans-National and Sub-National Cultures Another set of macro-level approaches, which considerably diverge from the those of the previous sets, are cultural studies that focus on trans-national and sub-national communities. For example, Bellah (2011) theorized that evolutionary biology developed communal capacities, such as dancing and storytelling, that fostered the ability for humans to participate in religious activities. Another version of conceptualizing religiosity as heritage is exemplified by Astor et al. (2017) in a study of Spain, and how religiosity facilitates a cultural historical relationship between religion and the state. Cichelli et al. (2020) identify global culture, including religiosity, to be part of larger processes to express personal and national preferences and lifestyle choices. Similarly, Karpov et al. (2012) investigate the intersection between religious and ethnic identities. Other approaches find Historical theories, such as the secularization thesis (e.g., Norris and Inglehart 2011;Hjelm 2018), were also focused on macro-level analysis in postulating that the world was becoming less religious overall. Though this theory has since been challenged as highly Western-focused and Christian-centric (e.g., Casanova 2011;Esposito 2010;Berger 1999), narrowly scoped on the exceptional cases of the United States and Western Europe, counter-evidence from macro-level data shows continued global religious affiliation (e.g., Conway and Spruyt 2018;Keysar 2017;Stonawski et al. 2015;Stark and Finke 2000;Finke and Bader 2017;Grim and Finke 2010). Additional examples of macro-level approaches include: Studies of overall demographic trends, such as the relationship between infant mortality rates and religiosity (e.g., Verona et al. 2010); religiosity and gender (Iyer and Joshi 2013); and religiosity and violence (e.g., Svensson 2016; Barter and Zatkin-Osburn 2014;Basedau et al. 2011).

Political and Economic Geographies
Another macro-level approach attends to national political and economic factors. These studies typically focus on the intersection of religiosity with the social institutions of governments or national economies (e.g., GDP). In some of these studies, the religious measures are considered to be outcomes (dependent variables) of political or economic causes (independent variables). For example, some economics-focused studies investigate whether monetary resources affect state regulation of religiosity (e.g., Müller et al. 2014;Buckley and Mantilla 2013;Wright and Palmer 2018), others whether economic insecurity fosters the need for religious coping and thus religion decreases as financial stability increases (Norris and Inglehart 2011). Another economics and religion approach is to study religious markets, considering religiosity to be a product that is consumed among other available consumption options (Woods 2012;Hungerman 2011;Iannaccone and Bose 2011;Finke and Stark 2011). Similarly, politically focused studies investigate whether governmental factors shape religious outcomes, such as vicinity of a nation state to other countries that regulate religion (Mataic 2018). In other studies, the causal direction is reversed, such that religious factors are investigated as potential explanations for other social outcomes. For example, studies investigate whether religiosity influences economic performance (McCleary and Barro 2019; GPEI 2018, 2019); social trust (Glanville et al. 2013(Glanville et al. , 2016Paxton and Glanville 2015;Paxton et al. 2014); education and printing presses (Woodberry 2010(Woodberry , 2012Nikolova and Polansky 2020); welfare attitudes (Van Heuvelen 2014); and public goods provision (Warner et al. , 2018a(Warner et al. , 2018bKılınç and Warner 2015;Schäfer 2012).
Trans-National and Sub-National Cultures Another set of macro-level approaches, which considerably diverge from the those of the previous sets, are cultural studies that focus on trans-national and sub-national communities. For example, Bellah (2011) theorized that evolutionary biology developed communal capacities, such as dancing and storytelling, that fostered the ability for humans to participate in religious activities. Another version of conceptualizing religiosity as heritage is exemplified by Astor et al. (2017) in a study of Spain, and how religiosity facilitates a cultural historical relationship between religion and the state. Cichelli et al. (2020) identify global culture, including religiosity, to be part of larger processes to express personal and national preferences and lifestyle choices. Similarly, Karpov et al. (2012) investigate the intersection between religious and ethnic identities. Other approaches find counter-trends for broad religio-ethnic identities. For example, Lu and Gao (2018) contest the notion of 'Chineseness' based on the existence of important within-nation religious diversity.

Summary of Macro-Level Approaches
In summary, an important aspect of studying religiosity is considering relationships with broad social institutions, such as governments, economic systems, and culture. Though most of the data in macro-level studies are collected from individuals, the key distinction from micro-level studies is that macro-level studies aggregate individually collected data into larger units, such as countries. Examples of keywords for macro-level studies are: Religious demography; religious affiliation; the names of particular world religions (e.g., Islam); religion and the state; religious regulation; religious freedom; religion and violence; religion and politics; religion and economics; religious markets and competition; secularization; ethno-religiosity; and religious heritage.

Micro-Level Approaches
This section reviews micro-level approaches. Collins may have coined the term 'micro-sociology': to "observe everything as closely as possible, to watch the action in the moments when it is taking place" (Collins 2010, p. 1). Applied to religiosity, this includes attending to phenomenon such as what else people are doing when they are praying, when last rites are administered to those who are dying, when a Buddhist monk has an experience of enlightenment, how and why someone is born again, and seeing the more general social patterns in these individual experiences. Through these examples, it is evident that Collins focuses on religious practices, and he refers to his and other's approaches, such as Durkheim and Goffman, as part of interaction ritual theory. Rituals, and religious practices more generally, are certainly an important aspect of interdisciplinary micro-level studies. Yet, religious practices are not the only form of individual-level religiosity and spirituality.
For example, Hackett (2014) reviewed survey measures of religious identity and found that how survey respondents self-identified did not always align with how scholars classified religious affiliation.
In particular, open-ended questions often yielded different results from closed-ended response options, indicating that adherents may have a syncretic religious affiliation that is not well-represented in mutually exclusive options that force respondents to select only one affiliation. Moreover, the analysis revealed the need to distinguish religious identity from belief and practice, which are not synonymous for all respondents. Most importantly, many questions about religious affiliation are asked in a way that presumes this is a static construct, but neither religious affiliation nor commitment were stable. For example, in analyzing data from the Panel Study on American Religion and Ethnicity (PS-ARE), Emerson et al. (2010) highlighted that panel studies can identify changes over time in how the same people answer questions about their religiosity and spirituality. The PS-ARE employs a growth design: Children of the respondents were also added to the sample once 18 years old. This approach facilitated the study of intergenerational transmission of religiosity.
To advance beyond religious affiliation alone, Saroglou's (2011) framing of the 'Big Four Religious Dimensions' helps to better address the full range of micro-level measures. The four dimensions are: (1) Belonging; (2) behaving; (3) believing; and (4) bonding. These four dimensions, while distinct analytically, are interconnected in reality. Moreover, the primary advantage of this approach is that the dimensions have applicability across various religious and cultural contexts and facilitate study of cultural variability in religion. Thus, this framework is employed in categorizing extant scholarship within the following sub-sections: Belonging, behaving, believing, and bonding. Plus, a fifth section attends to religious salience and spiritual identities, and a sixth to combinations.

Belonging
One of the most common ways to study micro-level religiosity is through the religious tradition affiliation questions that also inform macro-level studies. Globally, affiliation questions typically center around parsing adherents of Christianity, Islam, Buddhism, Judaism, Catholicism, and others. Within the United States, affiliation is often considerably more nuanced for within-Christian denominational groupings, which have been found to matter for a range of political, social, educational, and cultural distinctions. These approaches attempt to group the wide array of denominational and congregational types into larger categories, such as Evangelical Protestant, Mainline Protestant, and-due to the historical significance of the civil rights movement-Black Protestant. The relatively modern approach was formulated by Smith (1990), who drew upon prior classification schemes that focused on a fundamentalist-liberal continuum in U.S. religious traditions.
Upon this foundation, Steensland et al. (2000) offered a new classification scheme that adjusted for some of the factors not originally accounted for, as well as changes since the original schema was developed. The most significant change was attention to increasing nondenominational churches in the U.S., and attention to how to parse these into larger groupings of Evangelical versus Mainline Protestant. The resulting RELTRAD code groups congregations within four large categories: (1) Black Protestant (e.g., African Methodist Episcopal, National Baptist Convention of America, United Holiness, Zion Union, and more); (2) Evangelical Protestant (e.g., American Baptist Association, Lutheran Church-Missouri Synod, Wisconsin Evangelical Lutheran Synod, Wesleyan Methodist, Southern Baptist Convention, and more); (3) Mainline Protestant (e.g., Episcopal Church, Lutheran Church in America, United Presbyterian Church in the USA, United Methodist Church, American Baptist Churches in the USA, and more); and (4) Other Affiliation (e.g., Latter Day Saints-Mormons, Christian Scientist, Unitarian Universalist, and more). More than a decade later, Woodberry et al. (2012) updated this same larger categorization schema with a new list of congregational groupings.
A more contemporary approach emerged a few years ago: Lehman and Sherkat (2018) developed a religious classification that prioritized religious ideology, and deprioritized a separate category for Black Protestants, since this grouping can conceal the religious diversity present within many of the historically black denominations. There are seven Religious Identification categories: (1) Sectarian Protestants (e.g., conservative Protestant traditions, such as Baptists, Seventh Day Adventists, and Jehovah's Witnesses); (2) Moderate Protestants (e.g., Lutherans, Methodists, nondenominational); (3) Liberal Protestants (e.g., Episcopalians, Presbyterians); (4) Catholics; (5) Other Religions; (6) Mormon; and (7) No Identification. Shelton (2018) corroborated this schema as successfully differentiating meaningful group patterns in church attendance, prayer, conservative beliefs about the Bible, political ideology, sexual morality, and opposition to abortion.
In addition to these Christian-focused approaches, several studies also attend to micro-level religious affiliation with other major world traditions. For example, Chiswick (2014) studied Judaism in the United States, focusing on the ways that economics shape religious change over time. Though not complicating the measurement of religious affiliation with Judaism, these findings raise questions as to the degree to which affiliation is a constant, even with the same religious tradition in different times. Additionally, religious affiliation is studied in its intersection with ethnic identities. For example, Calvillo and Bailey (2015) found that Latinos in the U.S. have been shifting from historical roots in Catholicism to more Protestantism. To parse the significance of this religious tradition change within an ethnic population, the key measure of interest was language, specifically speaking Spanish at home. In conjunction with affiliation, this study revealed that homeland language use was more common among Latino Catholics than Latino Protestants. The implication is that Spanish-speaking, Latino Catholicism was a marker of ongoing emphasis on affiliation with country of origin, whereas English-speaking, Latino Protestantism was a marker of inculturation with the host country, i.e., the U.S. Moreover, when asked about their most salient identity, Catholics were more likely than Protestants to indicate their national origin or pan-ethnic identity. In comparison, Protestants were more likely to name either their religious or American identity.
Alongside measurements of religious affiliation are studies that attend to the lack of affiliation with religious traditions, or the so-called religious "nones." For example, Vargas (2012) reported that nearly one-fifth of Americans were religiously unaffiliated. Based on data from the Portraits of American Life Study (PALS), the results revealed that an even greater proportion gave serious thought to leaving religion, and the distinctions were examined between those who actually disaffiliated, versus those who thought about leaving but remained affiliated. Studying the United States and Canada, Thiessen and Thiessen and Wilkins-Laflamme (2017) examined the extent to which religious nones were socialized to adopt a "no religion" position as children, as compared with disaffiliating during their teenage or adult years. Religious nones in Alberta, Canada provided the main qualitative sample, and Alberta had a slightly higher proportion of religious nones versus Canadian or American national averages. The in-depth interview findings indicated that one of the primary reasons for rising disaffiliation among young people was that parents were increasingly allowing their children to decide whether to continue their religious affiliation. Another major reason was young people having intellectual disagreements over core issues that slowly developed doubts about ongoing affiliation. Young people also disaffiliated because their friends left, and because of significant life transitions in their home, such as changing jobs, getting divorced, leaving to attend a university, or losing a parent or grandparent who was the main influence on the family's religiosity.
More generally, several of these micro-level approaches to studying religious affiliation revealed that this is not a static category and instead can be dynamic over time as life circumstances change. For example, Bakker and Paris (2013) found that major life events, in particular the loss of a child during childbirth, was related to fluctuations in religiosity. Analyzing participation in online bereavement support groups, the results revealed that some grievers experienced religious disorientation, expressed anger at transcendent reality, questioned their belief systems, or left their faith altogether. On the other hand, other sets of grievers experienced religious reorientation, deepened religious conviction, resumed affiliation, or in other ways renewed their spiritual interests. This last study, in particular, highlights the intersections between belonging and behaving.

Behaving
The second most prevalent micro-level approach to studying religiosity is behaving, such as religious service attendance, prayer, reading religious texts, and ritual. One of the most common ways to measure religious practice is by asking about religious service attendance frequency. In surveys, this question is fairly standardized to: "Aside from weddings and funerals, how often do you attend religious services: more than once a week, once a week, once or twice a month, a few times a year, seldom or never?" Frequency of religious attendance is predictive of a wide array of social life. For example, Schwadel and Falci (2012) found that attendance, coupled with religious affiliation, parsed who had more or less depressive symptoms. Specifically, evangelical Protestants who attended frequently had more depressive symptoms than regularly attending Catholics. Among those who attended services less frequently, evangelicals reported lower levels of depression than mainline Protestants who attended less frequently. It is worth noting that for Nebraskans as a whole, church attendance did not have a significant relationship with mental health. Importantly, it remains unclear from this study which way the causal direction occurs in this relationship, as it could be the case that people self-select into different denominations based on their wellbeing, which is also correlated with other important factors, such as income and educational attainment. Nevertheless, this study indicated that affiliation alone was not sufficient to understand mental health, but affiliation in conjunction with attendance frequency was indicative. In a second example of relationships with attendance, Burdette et al. (2018) found that, among urban mothers in the U.S., more frequent religious attendance was related to reduced use of illicit drugs. This was interpreted as relating to religious attendance reinforcing regular social contact with other church members.
Another prevalent measure for religious behavior is prayer. A typical approach to studying prayer, as with service attendance, is to inquire about frequency. For example: "Aside from religious services, do you pray several times a day, once a day, a few times a week, once a week, a few times a month, seldom, or never?" Another approach is to ask about prayers during meals, participation in prayer groups, and fasting. For example, in a study of U.S. college students, Ozorak (2003) found prayer to be meaningful in parsing who was most likely to volunteer again after a campus service-learning experience. In this case, prayer was not measured as a frequency, but rather through a prayer scale that asked about experiences with God when praying, such as: "I feel very close to God when I pray", "Most of my prayers are for God to solve problems for me", "Confessing my sins to God helps me live a better life", "I often ask God to strengthen me so that I may help others", "When I pray, I feel secure", "I often pray on behalf of other people", "When God has answered my prayers, I always give thanks", and "When I pray alone, I have a ritual that I adhere to strictly" (Ozorak 2003, p. 289).
Sometimes studies also focus on ritual. For example, Orsi (2013Orsi ( , 2018 studied Catholic rituals surrounding the Virgin Mary and the communion host. Additionally, Baker (2010) applied Collins' interactive ritual framework in studying collective rituals, finding that the most effective in providing 'religious goods' were those that were full of emotional energy, satisfaction, and group belonging. When experiencing this combination of ritual experiences, members were more likely to retain commitment to the religious group, resulting in stronger religious tradition affiliation (see meso section for more on this). More generally, Ammerman (2014Ammerman ( , 2015 referred to this approach to studying religious behaviors as attending to 'lived religion,' or everyday practice. Criticizing quantitative approaches for too readily focusing only on what can be easily counted, Ammerman stated that attending to religious and spiritual practices rendered the invisible observable, especially aiding a centering on the lives of women, people of color, and people in the Global South. Believing A third way of studying micro-level religiosity is investigating beliefs. Beliefs are more difficult to measure in a uniform way, across different religious traditions and diverse cultures. For example, a common approach (in Western contexts) is to ask if one believes in God, and then follow that with a series of questions designed to elicit beliefs about God (e.g., God rewards the faithful, God punishes sinners, I have a warm relationship with God). While the term God can be translated to multiple languages and can be replaced with alternative terms, such as Allah, the phrasing of the question may not be suitable for all forms of religious beliefs (e.g., Buddhism). Yet, beliefs can be a deeper way of understanding religiosity, beyond the easier to count measures of service attendance and prayer frequency. Plus, belief measures can reveal heterogeneity within people who all select the same religious affiliation on closed-end survey questions. For example, among people selecting Catholic religious affiliation, there is tremendous variation in whether people believe that the host is an instantiation of the incarnate, not merely a symbol (e.g., Smith and Snell 2009). Yet, in formal Church doctrine, this is perhaps one of the most defining features of Catholic beliefs. Also, Catholics vary in the extent to which they believe that attending mass regularly is important, when the Catholic Church states that it is a requirement, and more examples, all which illustrate that affiliation is not synonymous with individuals holding the beliefs that the official denomination indicates that they would. To complicate matters further, Wuthnow (2010a) found that people can hold a collection of beliefs garnered from different world religions, some of which are contradictory. For example, many younger Americans affiliate with Christian denominations while believing in Kharma, albeit typically by thinking of it as 'karma' and not as situated within a set of plural-theistic Hindu beliefs.
Additional examples of belief measurement include concepts such as duty, love, transcendent, involved being, and energy force. Other beliefs can be about religious texts, such as whether the Bible is believed to be literal or not. For example, Hempel et al. (2012) found that a personal commitment to Protestant theological conservatism is negatively associated with the likelihood of trusting strangers. This conservatism suppressed the positive relationship between religious attendance and generalized trust among Christians. Further examples of beliefs include specific interpretations of larger beliefs. For example, American Christians commonly believe that good things come from their faith. This general belief is distinct from a particular emphasis on 'prosperity gospel': The belief that God explicitly rewards the faithful. For example, Dougherty et al. (2019) examined the impact of prosperity beliefs on entrepreneurial activity in the United States and found that prosperity beliefs moderated the impact of entrepreneurial values on the likelihood of starting a new business venture, but such beliefs themselves did not transform individuals into entrepreneurs. While prosperity beliefs were associated with self-enhancement attitudes, they also were associated with a reduced openness to risk, which was needed for entrepreneurial activity. Notably, researchers removed religious tradition and self-rated religiosity from the model because these variables tended to be inconsequential. In other words, this moderating relationship of prosperity beliefs to values, entrepreneurialism, and business creation was not confined to any single religious denomination.
In another example of the study of beliefs, Warner and colleagues first qualitatively investigated the content of religious beliefs among Muslims and Catholics, in countries where those religions were the majority (Warner et al. , 2018a(Warner et al. , 2018bKılınç and Warner 2015;Cohen et al. 2017). After identifying duty to God or Allah as a belief that was seemingly held in common across both religious groups, but with important variations, they then designed an experiment to assess what effect priming religious expressions of duty to God had, if any, on each group's election to donate to a charitable cause, as compared to a generic prime in the control group. Specifically, the qualitative phase of the mixed-methods study revealed that both Catholicism and Islam valued belief in duty to God but in distinct ways. Islam emphasized obedience as key to this duty, whereas Catholicism emphasized adherence to the commandments, but with a greater degree of indifference, at least for the lived experiences of many Catholics, to actually enacting duty in any stronger way beyond the parameters of not killing, cheating, or coveting too much, and trying to be nice to neighbors. With these distinctions in mind, the experimental phase of the study was designed to test the hypothesis that duty to God should work differently in these two groups, despite on the surface seemingly being a similar belief. The evidence supported a difference in the way the belief activated behavior, at least insofar as the activation primed the two groups to give charitably in distinct way. Counter-intuitively, the more intense duty activation for Muslims resulted in less giving, with the theory being that the Muslims may have already been reaching the height of their obligation to give before the study and thus could not be primed to give any more. Alternatively, the more indifferent version of Catholic duty appeared to activate Catholics to give more, since the reminder of the duty, that was otherwise latent and not yet fully maximized, resulted in greater observance of the belief in action.
Religious beliefs are particularly important within the study of health and wellbeing. Indeed, a robust subset attends to measures of religious coping (e.g., Konieczny et al. 2012;Ai et al. 2013). For example, religious coping has been investigated within studies of grief and death anxiety (Feldman et al. 2016); health (DeAngelis et al. 2019); illness recovery (Testoni et al. 2016); mental health in mothers with intellectually disabled children (Sharak et al. 2017); genocide survivors (Fox 2012); and sexual violence (Cranney 2017). Of particular note within this approach to studying religiosity is the religious coping scale (Pargament et al. 2011). Since this approach studies relational features of transcendent beliefs, it and similar approaches are reviewed in the next section: Bonding.

Bonding
These bonding approaches can be considered a subset to beliefs, since many of the relational approaches to studying micro-level religiosity focus on a person's relationship to God. Notably, the wording in many of these approaches is limited to God, since most of them were developed in Western, Christian-majority contexts. Nevertheless, several of the studies have been investigated outside of the U.S. and Western Europe, and there are insights about studying religiosity regardless. For example, developed in the 1990s and then refined to a briefer, 14-item scale, the RCOPE is designed to measure the ways that people rely upon their religiosity in responding to trauma, crisis, and other transitions (Pargament et al. 2011). This is not meant to imply that people only rely upon their religiosity in positive ways, as the scale codes 'positive religious coping' to be ways that people turn to their religious relationship securely and out of a sense of spiritual connectedness. Alternatively, 'negative religious coping' refers to people turning to their religious relationship in troubling ways, which the researchers describe as reflecting: "underlying spiritual tensions and struggles with oneself, with others, and with the divine" (Pargament et al. 2011, p. 51). All of these religious coping measures are predicated on a belief in God, and seemingly a personal, interventionist God (more on this below), which is why this approach is included here, in the bonding-belief section. Specifically, the short-form RCOPE scale includes two groupings of 14 items (see Appendix A).
On the one hand, this scale could be, and has shown to be, a powerful predictor of different coping styles. On the other hand, when considered alongside the belonging and other belief measures above, it is also readily evident that the scale is likely to be tapping several of the religious affiliation and theologically relevant belief measures. Particularly concerning is the way that several of the characteristics of sectarian Protestant faiths (Lehman and Sherkat 2018;Shelton 2018) are coded as negative religious coping. This is especially relevant in considering that-though religious affiliation is analyzed alongside the coping measure-Protestant faiths are all collapsed and reported among the following affiliation categorization: "thirty-three percent of the sample was reported as Protestant, 22% as Catholic, 2% as Muslim, 1.5% as Jewish, and 2% as having no religious affiliation" (Pargament et al. 2011, p. 58). Yet, this is within a meta-analysis reviewing 30 studies employing the RCOPE scale, in which several appear to have disaggregated Protestantism.
Another example of micro-level religious bonding is from Manglos-Weber et al. (2016). Studying a nationally representative sample of young Americans, the researchers investigated relationships with God. Of particular interest were four divine bonding facets: Intimacy, consistency, anxiety, and anger. Each of these four facets was asked about through three items each, resulting in 12 total items (see Appendix A). Other approaches to studies of bonding include awe and forgiveness. For example, Van Cappellen and Saroglou (2012) conducted experiments to assess whether inducing awe affected feelings and behaviors. They defined awe as "a feeling of wonder experienced by the self when facing something vaster, greater, beyond current understanding" (2012, p. 225). Examples included childbirth, panoramic views, and great works of art. Feelings of awe were studied in relation to other measures of religiosity and spirituality, as well as with experimental outcomes, such as travel decisions. Similarly, Piff et al. (2015) described awe as a collective emotion that is: "an emotional response to perceptually vast stimuli that transcend current frames of reference" (2015, p. 883). Within the combination of two experiments, the researchers found that tendency for awe predicted greater generosity, and that inducing awe increased generosity, ethical decision-making, and prosocial values. Additionally,  studied the relationship between awe of God and participation in religious congregations. While the congregational aspect of this study is included in the meso-level section below, the awe of God measures are situated here in the micro-level section. Awe of God was asked about through six items (see Appendix A).
Lastly, Webb et al. (2005) and Thompson et al. (2005aThompson et al. ( , 2005b studied the ability for forgiveness in relation to religious faith. In a Christian, U.S. sample, college students answered surveys about personal God beliefs, religious behaviors, and forgiveness. Forgiveness was measured through the Heartland Forgiveness Scale and included three subscales: Forgiveness of self, others, and a situation (e.g., natural disaster). All these facets shared in common a person's ability to disregard a tendency to resent negative life circumstances in forgiving the perceived source of harm. The results indicated that people with greater religious practice were more likely to forgive, as were people with a belief in a loving deity, and those who employed religious coping to solve problems. More recent studies of forgiveness include Jang et al. (2018), who asked respondents their frequency in doing the following: 'I have forgiven those who hurt me.' and a study in Zimbabwe by Tarusarira (2020) that investigated complex dynamics surrounding forgiveness as related to peace, social justice, and reconciliation, finding that truth-telling is often a necessary pre-condition of community forgiveness.
In summary, relational bonding aspects of religiosity, between an individual and a divine being, are an innovative aspect of micro-level studies. Yet, it is worth noting that these approaches are highly contingent on belief in a personal form of transcendent being that best characterizes Christianity, perhaps specifically evangelical Protestantism. For example, Manglos-Weber and colleagues stated: "This work and the following conceptualizations primarily apply to the majority of young adults who share cultural frameworks rooted in American Christianity, and much of the research on relationships with God focuses on evangelical Christianity . . . The applicability of these conceptualizations to non-Christian faiths requires further study" (Manglos-Weber et al. 2016, p. 4).

Spiritual Identities and Religious Salience
A fifth approach to studying micro-level religiosity concerns spiritual identities, also sometimes referred to as religious salience or importance of faith. These approaches attend to religiosity as one among many identities: Viewing people as holding many identities and affiliations in tandem, but not all with the same degree of emphasis. While two people may have affiliation with the same religious tradition, one person may place a great deal of emphasis on this religious affiliation, while the other may not see their affiliation as a key aspect of their identity. In delineating this, Curtis and Olson (2019) examined religious identification in two ways: (1) Importance: Strength of attachment, and (2) prominence: Prioritization of religious identity among other identities. Specifically, they asked respondents: "We are all part of different groups. Some groups are more important to us than others when we think of ourselves. How important are each of the following in describing how you personally see yourself?" Respondents were then asked to rate their strength of identification with 14 separate identity items (including religion; see Appendix A for full list; Curtis 2013). Studying Germany, Poland, and the United Kingdom, the results indicated that attachment to religion was variable cross-nationally. As for prominence, while religious identification was slightly higher among Polish respondents, religion was not a salient identification for most of these Europeans. Based on these results, the researchers cautioned that a more nuanced understanding of religiosity needs to include a diverse geographic sample and a wider array of measures than affiliation alone. Similarly, Gebauer et al. (2012) measured religious salience by asking respondents how much they agreed or disagreed with the statement: "My personal religious beliefs are important to me." Similarly, the National Study of Youth and Religion in the U.S. asked teens and young adults: "How important is your religious faith in providing guidance in your own day-to-day living?" with a range between extremely important to not important at all (see Appendix A ;Smith 2003;Smith and Snell 2009).
In addition, many people throughout the world hold spiritual identities that are not necessarily religious. For example, Bredle et al. (2011) studied the Functional Assessment of Chronic Illness Therapy scale (FACIT), which has a subscale containing 12 spiritual indicators (see Appendix A). These items are then scored into three constructs: Meaning, peace, and faith. The existence of spiritual identities led scholars to investigate the notion of spiritual but not religious. This is largely an American term that Ammerman (2013) described as originating from a quantitative approach that studied religiosity as one construct, and spirituality as another, conceiving of essentially a two-by-two table in which one could be high or low on each factor, and thus result in a spiritual-but-not-religious (SBNR) category that is high on spirituality and low on religiosity. Qualitative analysis instead revealed at least four ways that Americans held spirituality and religiosity in tandem.
More generally, studies such as these indicate that spirituality is not collapsible with religiosity and deserves separate, or joint, attention. Indeed, Steensland et al. (2018) described this within the U.S. as a result of an increasing trend toward informal spiritual seeking and away from formal religious affiliation. Building upon several existing sets of measures, the researchers delineated 13 dimensions of spirituality based on referent categories (such as referent to God or Jesus, a higher power or a supreme being, and the self; see Appendix A). Additionally, six spiritual orientations described the type of connection had with the referent: Relational, emotional, cognitive, ethical, behavioral, or existential. For example, cognitive orientations to a divine being included referring to connections with a higher power in terms of knowledge or belief, whereas relational orientations to a monotheistic deity included referring to connections with Jesus as a personal relationship. Combined, these measures resulted in a set of seven classes that described the contours of various types of spirituality: (1) Spirituality as organized religion, (2) orientations toward God: belief in God, (3) orientations toward God: Relationship with God, (4) beliefs outside particular traditions: Belief in a higher being, (5) beliefs outside particular traditions: Belief in something beyond, (6) relational spirituality: Holistic connection, and (7) ethical action. The researchers analyzed how these different spiritual classes related to other social factors, such as age, education, gender, race, and ethnicity.

Combinational Micro-Level Approaches
Since many scholars recognize the importance of studying micro-level religiosity through multiple measures, there exist several scales that combine indicators of religious belonging with behaviors, or behaviors with beliefs, or belonging with religious salience. A historically important approach, with an ongoing set of measures in contemporary studies, was defined by Allport and refined with Ross (Allport and Ross 1967): The intrinsic and extrinsic religiosity scale (Kirkpatrick and Hood 1990). The scale builds from self-determination theory, which views intrinsic motivations as orientations or behaviors sought without any extrinsic rewards, returning only internal satisfaction and fulfillment (Neyrinck et al. 2010). With this theory as context, Allport and Ross sought to differentiate religious behaviors that are externally required or reinforced from personal, spiritual, internal desires for religiosity. The original measures were validated by Hoge (1972), revised and formulated into single-item scales by Gorsuch and McPherson (1989), and employed within several studies since (e.g., Cohen et al. 2017). The result is a 20-item scale with two subscales distinguishing extrinsic from intrinsic religiosity (see Appendix A). Notably, the original scales were formulated within U.S. samples, for example, Gorsuch and McPherson conducted their study within a sample of Californians. As with some of the bonding measures above, there are implicit Christianity beliefs embedded in the way the items are asked (e.g., God's presence). Nevertheless, this scale exemplifies how scholars combine multiple aspects of micro-level religiosity (belonging, behavior, believing, and salience). Indeed, Batson (1976) described this combinational approach as confounding multiple factors: Religion as a quest versus the 'true believer of the intrinsically religious individual' (Batson 1976, p. 32). Batson thus employed the same measures in a study with New Jersey Princeton seminarians to develop a quest scale that parsed the IR scale into three components using factor analysis: Religion as means, religion as ends, and religion as quest. He then studied how each factor related to prosocial helping.
Additional iterations of the IR scale include the Spiritual Experience Index (SEI) developed by Genia (1997), which combined Allport's formulation with items developed from other major approaches (e.g., Mature Faith Theory) into 11 criteria, toward studying diverse spiritual affiliations and distinguishing those who are 'spiritually mature from those with less evolved forms of faith' (Genia 1997, p. 345). Despite the word experience in the index title (which implies a focus on behavior), these criteria intended to measure belonging, behavior, beliefs, and bonding (Genia 1997, p. 345). The result is a proposed paradigm to investigate these criteria within two factorial constructs with accompanying subscales: Spiritual support (SS) and spiritual openness (SO) (see Appendix A). This approach reveals, particularly within the openness subscale, an intention to downgrade belonging by reverse scoring affiliation with a particular religious tradition (especially focused on particular sectarian Protestant faiths, it seems, in item 10 of the openness scale), while combining measures of behaviors, beliefs, bonding, and salience. In other words, people are ranked high on the openness scale if they do not affiliate strongly with particular religious traditions but are still interested in religions generally, which is labeled 'religion as quest.' Notably, this too was a U.S. study, though the particular within-U.S. locations are not reported.
Continuing a similar combinational approach to studying religiosity, Koenig and Büssing (2010) further developed the IR scale from Allport, Ross, and Hoge. Along with insights garnered from a conference on Methodological Approaches to the Study of Religion, Aging, and Health, hosted by the National Institute on Aging and the Fetzer Institute, the researchers focused on measuring two facets of religiosity: Organizational religious activity (ORA) and non-organizational religious activity (NORA). The former included formal religious behaviors such as attending religious services, while the latter distinguished individual religious or spiritual behaviors, such as praying or reading religious texts (referred to as Scripture). Though the labeling of theses subscales focuses on social versus individual behaviors (measured with the first two items, both as frequencies), the other three items also attend to belief, bonding, and salience, and are denoted as IR (Koenig and Büssing 2010; see Appendix A). Notably, Koening and Büssing report that "the DUREL was designed to measure religiosity in Western religions (e.g., Christianity, Judaism and Islam), and may be less accurate in its assessment of religiosity in Eastern religious traditions (e.g., Hinduism or Buddhism). However, this does not necessarily mean that the DUREL is not valid in these populations, but only that the wording may need to be adapted to those specific religious traditions. For example, the word 'church' could be replaced by 'temple' or 'mosque' for non-Christian samples. Likewise, the word 'Bible' could be replaced by 'Torah' or 'Koran' or 'writings of Buddha.' Further studies in these populations are needed to confirm or refute this claim, including validation using changes in wording" (2010, p. 84). To date, one such replication study exists: Lee and Baumann (2019) investigated DUREL in South Korea, among psychiatric staff in the city of Daeugu and suburbs of Chilgok and Yeongcheon.
Another combinational approach was developed by Huber and Huber (2012). Their scale builds upon a framework from Glock (1962Glock ( , 1973) that focused on five core dimensions of religion, the intellectual, ideological, ritualistic, experiential, and consequential. The framework was revised by Stark and Glock (1970) to drop the consequential dimension and to split the ritualistic dimension into public versus private practice, thus resulting in a remaining total of five dimensions. Huber and Huber recognized a focus in extant measures on North American Christianity, and thus the next several decades involved revisions to the items to aid their generalizability beyond Western geographies.
Huber then offered the Centrality of Religiosity Scale (CRS) as an improved approach that was intended to be more broadly geographically applicable. The scale included at least three variations, a 15-item version, a 10-item version, and a 5-item version. Additionally, several items were developed to measure 'interreligious' samples, included within interreligious scales in various sizes: 7-item, 14-item, and 20-item versions (Huber and Huber 2012; see Appendix A). Notably, the original items were developed in German and then translated to 19 other languages. The researchers stated that "these versions suitable at least for Abrahamitic religions (Judaism, Christianity, Islam) are nested in each other and grow more economical" (Huber and Huber 2012, p. 716).

Summary of Micro-Level Approaches
In summary, there are a plethora of micro-level approaches to the study of religiosity, which include measures of belonging (e.g., affiliation), behaving (e.g., service attendance and prayer), believing (e.g., universal energy, transcendent and personal divine being), bonding (e.g., emotive qualities of a relationship with a divine being, attachment, experiences of awe, forgiveness), salience (e.g., importance of faith, affiliation, particular beliefs in everyday life and among other identities), and spiritual identities that are not necessarily associated with formal religiosity (e.g., peace and harmony with all of humanity). In addition, a number of combinational approaches exist that merge several micro-level indicators into the same scale, sometimes even within the same measure.
Indeed, scales are particularly common within micro-level studies, such that scale could be a helpful keyword for parsing studies by level. Micro-level approaches to studying religiosity and spirituality are particularly important in understanding the intersection with health and wellbeing, such that the keywords health and wellbeing, along with religiosity and spirituality, are another way to initially identify micro-level studies. Additional micro-level keywords include: Faith, prayer, religious or spiritual practice, belief, ritual, religious coping, relationship with God, awe, forgiveness, religious or spiritual salience, importance of faith, intrinsic or extrinsic religiosity, public or private faith, and spirituality (also: Spiritual with identity, seeking, openness, experience, or support). Religious affiliation, tradition, denomination, and service attendance are also keywords that are employed in micro-level studies, but since these are engaged regularly in macro-level studies too, further parsing is needed to identify the level.
Reviewing macro and micro-level approaches together reveals that micro-level indicators, on the whole, are considerably more tailored to Christian-specific indicators, especially for extant measures of beliefs and bonding. The relatively narrow focus on affiliation within macro-level studies facilitates broader applicability in cross-national, multi-national, and single nation studies outside of the United States and Western Europe (Herzog 2020). Yet, the greater depth in micro-level approaches for multiple facets of religiosity and spirituality raises significant questions regarding the extent to which affiliation alone is a valid indicator of religiosity.
At the least, the salience approaches in micro-level studies indicate the need to also measure the degree of importance for affiliation. In reverse, the macro-level studies condition the generalizability of several micro-level approaches, as the smaller proportions of Christians in global geographies limits claims that belief and bonding measures, in particular, can simply be translated and slightly reworded to apply to interreligious samples.
Relationship with a personal divine being requires more attention than wordsmithing. Indeed, since Finke (2014) stated that macro-level approaches are typically built on micro-level theories, micro-level studies should be attuned to the global limitations of measures. More generally, desires to limit Christian-centrism need to carefully consider whether the goal is to develop measures that can be standardized across all contexts and cultures, or whether the goal is to develop context-specific, culturally appropriate measures that can be aggregated in thoughtful ways. Additionally, Finke (2014) highlighted the relative dearth of meso-level studies, which are reviewed next.

Meso-Level Approaches
This section reviews meso-level approaches to studying religiosity and spirituality. Much of religiosity and spirituality happens through group interaction such as in families, local communities, and global networks, as well as through religious and faith-based organizations. Meso-level approaches range from a focus on informal networks and groups, to occupations, to formal organizations. Within formal organizations, a significant proportion of studies attend to congregations and denominations as the site of data collection. Other meso-level approaches focus on faith-based organizations that are not explicitly congregations. Admittedly, the boundaries between where micro ends and meso begins can get blurry, as is true of where meso ends and macro begins. Figure 3 (above) aids an analytically clean distinction between the levels. In reality, the levels are best understood not as separate 'buckets' that are entirely mutually exclusive, but rather as a spectrum that gradually shifts from one level to the next. Metaphorically, envision a semi-circular thermometer, as can be employed for fireplaces or stoves. A yellow color on the left designates micro-level approaches, and the red color on the right designates macro-level approaches. The blended color of orange in the middle signifies that meso-level studies are an intersection between micro and macro levels. Moreover, an incremental, gradational change in hues between yellow-orange and orange-red represent the approaches that lie at the porous boundaries that connect each level (e.g., micro-meso).

Religiosity and Social Networks
Religious networks are studied at the juncture between micro-level and meso-level approaches and often focus on how participation in informal social groups fosters interpersonal experiences, with individual inputs and outcomes. This includes individual experiences of social support, or exclusion. It also includes the relational ties formed, strengthened, or broken within religious contexts (Merino 2014). For example, Wuthnow (2010b) described one of the central roles that religiosity has in social life as being a site of what he called 'intimate knowledge': Religious networks and communities are a microcosm of public space that fosters interpersonal connections around individual spirituality. In this way, religious networks can be viewed as similar to the extended relational ties at a family reunion, bound in certain ways around knowledge of each other that is less than completely public, and yet also diffuse and with the ability for distant relational ties to not know each other that well, or remember each other's name, despite also having access to more personal knowledge of one another.
The most informal and micro-level-leaning approaches within meso study are those that focus on interpersonal relationship dynamics. At the most basic level, a social group is constituted by more than one person, meaning a dyad is the smallest social group unit. Relationships are highly dependent on the ability of one person to understand the perspective of another, which is often referred to as empathy. Some approaches consider empathy to be a fixed quality, akin to a personality characteristic, but most view empathy to be malleable: A quality that is changeable and responsive to social stimuli (e.g., Schumann et al. 2014). In this sense, empathy is an action word, an effort that people engage (or do not) in support of relational understanding. Studies of empathy find that the degree to which it is present at any particular point in time, or within certain groups, is variable. For example, Konrath et al. (2011) found that dispositional empathy has been decreasing over time in American college students. Among the subscales for empathy, empathic concern and perspective taking were especially low and decreasing in supply among this group since 2000. Related to religiosity, Trothen (2016) suggested that the presence of religious pluralism can develop empathy, at least if it is accompanied with concerted efforts to aid members of diverse religious groups to understand and value the views of members from other religious groups.
Additionally, one reason that the behaviors studied within micro-level approaches may matter, in terms of affecting social outcomes, is through group-level dynamics. For example, Fuist (2015) studied the effect of collective prayer, specifically talking to God within a group of witnesses. The results indicated that this group ritual was distinct from, and not simply an aggregate of, individual prayer. As a group process worthy of attention, collective prayer entails social performance, instances of empathy and solidarity within a larger group, exclusion, and other experiences that are characteristic of group participation generally. Indeed, Graham and Haidt (2010) stated that extant approaches have often treated religiosity "primarily as a set of beliefs held by individuals" and yet "beliefs are only a facet of this complex and multidimensional construct" (Graham and Haidt 2010, p. 140). The researchers thus studied religious practice as a means of creating 'moral community' in which sets of moral values were fostered and held in common among similarly perceived others. The study of groups generally, and religious groups specifically, indicates that communities create, revise, and are based upon a mutually shared set of norms: Inter-subjective understandings of what constitutes acceptable or undesirable behavior (e.g., Jeppe Sinding 2013). Within this context, there are several group-level properties that are important for understanding meso-level religious dynamics. For example, Hoverd et al. (2012) found that the size of groups matters. In New Zealand, members of smaller religious groups, which constituted less than 1.5 percent of the population, tended to strongly identify with their religious affiliation. In comparison, members of religious groups that constituted 6 percent or more of the population tended to have looser connections to their religious affiliation. This may be because larger religious groups are less cohesive and experience more contested identities and ideological differences. Regardless, this meso-level study of religious groups helps to explain why variability within religious salience (micro-level) can exist within multi-national data of adherents from the same religious tradition (macro-level).
Focusing on the group and community aspects of religiosity raises complex questions regarding how to define and measure what constitutes a religious group, as compared to non-religious social groups. For example, it may seem an oddity upon first exposure, but atheist communities are also a religious group, insofar as their collective identity is defined by a shared anti-religious belief in the certainty of the non-existence of God (e.g., LeDrew 2013), which is distinct from agnostics who are unsure about the existence, and varying degrees of certainty and doubt among religious adherents. Additionally, several studies identify yoga groups as culturally appropriated, religiously derived communities that embody many of the characteristics of other religious groups (e.g., Jain 2014, 2020). For example, Yi and Silver (2015) complicated simplistic narratives of a rise in religious nones (macro-level) by observing that most nones claimed to be neither atheist nor agnostic, rather they are mostly disaffiliated. As the micro-level section revealed, disaffiliation (one form of belonging) is not necessarily synonymous with an absence of belief, bonding, or behaving. "Rather, the decline of formal religious participation may reflect and reinforce a broadening sense of contingency of any given religious orientation and a proliferating variety of spiritual-ethical options (Taylor 2003)" (Yi and Silver 2015, pp. 599-600). Studies of new spiritual movements, including the prominence of yoga groups, reveal that the sets of religious and spiritual beliefs are more diffuse and less regulated by an organizing body, yet nevertheless characterize identities of participating individuals in shared ways. 1 One could even identify a set of uniform practices for participants, e.g., clearly identifiable clothing. For this reason, there is controversy over the presence of yoga practices within public schools in the U.S., which are required by law to not promote any particular religious tradition (Brown 2019).
In terms of outcomes of participating in religious and spiritual communities, these can be positive and negative, even both simultaneously between different sets of members, or within the same members over time. For example, numerous studies identify a relationship between participating in religious groups and having a higher degree of social trust. This could be due to religious groups facilitating social trust, or people with greater levels of social trust joining religious groups. In parsing this, Seymour et al. (2014) found that interpersonal processes within religious groups mattered, such that relationships with religious leaders was more predictive of trust levels than was religious beliefs or practices. Additionally, Dingemans and Ingen (2015) found that religiosity can be both 'trust-enhancing' and 'trust-reducing.' Studying these dynamics within the European Values Survey, the results indicated 1 Individually, yoga can intersect with religiosity in a variety of ways; the point here is a about yoga groups.
Religions 2020, 11, 437 22 of 74 that multi-level effects were important for understanding these complex relationships. Specifically, religious affiliation generally was not strongly predictive of trust levels. However, Protestantism was strongly correlated with greater levels of social trust, yet this was highly contingent on country location. Moreover, religious service frequency was positively related to greater trust, yet this was conditioned by the relationship between religious tradition and residual effects of childhood participation. When controlling for all these other facets (which may be 'absorbing' many of the important religious effects), bonding with God was negatively related to social trust. Notably, the findings varied by geographic context, such that greater religious diversity within nations associated with greater levels of trust, relative to nations with lower levels of religious diversity.
However, the study of religious diversity is itself complex. Wuthnow (2011) highlighted the challenges of religious diversity within the United States, in terms of doing the 'hard work' of religious pluralism, which he viewed to hold several opportunities (despite the book's title). The work referred to efforts to delve beneath superficial treatment of diverse religious groups that feign respect while not offering deep respect for irreconcilable distinctions between the beliefs of different religious traditions. In other words, true respect is not pretending that all groups can get along in a 'watered-down' middle-ground that amalgamates beliefs from multiple traditions. Pluralism does not equal a muddled homogenous composition of beliefs selected from many traditions, and some members of religious traditions may view such an approach to be highly disrespectful of, or ignoring, valued distinctions. For example, it is not respectful or accurate to presume that Christianity and Judaism can be combined into a homogenous Judeo-Christian, when one faith places heavy emphasis on the transcendental purposes of Jesus and the other views Jesus as one among many historical figures. That is an important distinction, and true pluralism is not ignoring or demeaning this importance. Rather, the 'hard work' of pluralism is to respect the existence of these divides, and yet to seek a broader sense of social solidarity that honors the existence of these irreconcilable beliefs.
Relatedly, a meso-level focus on the group dynamics of religiosity helps to reveal the in-group and out-group dynamics that surround affiliation with religious groups. Salvatore et al. (2020) refers to group participation as a condition within the salience of 'otherness': Simplification with contexts of tremendous diversity into relatively simple 'sense-making' categories of us and them. Boundaries of otherness are drawn to respond to a cognitive need to reduce complexity. That seems to be a basic human need, and yet a long-standing social problem is consistently and persistently drawing those boundaries along racial and ethnic lines. Religiosity is embedded within, and overlapping, racial and ethnic identities. Indeed, some scholars refer to the concept of religio-ethnic identities to acknowledge the inability to disentangle religiosity and ethnicity. With this context in mind, religious groups can be understood to be part of needed attention to ethnic diversity and in/out-group dynamics. For example, Jung (2012) studied the dynamics of Islamophobia in the U.S. as an in-group and out-group dynamic that is conditioned by the degree of inter-group contact, a common approach in studies that focus on race and ethnicity more generally. Additionally, Rytter (2014) studied the ways that the Sufi movement, within Islam, defines subcultural identities around the globe, within and crossing national boundaries. These studies complicate approaches that study only micro and macro levels.

Religiosity and Occupations
A similar but more formalized approach to studying group-level dynamics and religiosity is the study of occupations. While often diffuse and loosely coupled, occupations nevertheless provide organizing structures that condition, and are conditioned by, individuals and nations. Professional associations gather individuals across geo-locations and help to define a set of bounded norms for acceptable, common, and ethical actions at work. Beyond their boundedness within work contexts, some professional identities supersede work-only actions by defining broader lifestyle choices. For example, many medical professionals extend beyond discussing nutritional principles at work to enacting those actions into their broader lifestyle. Within this context, Ecklund and colleagues studied the intersection of religion and science by investigating what role religiosity had within scientific occupations (e.g., Ecklund 2010; Ecklund and Lee 2011;Ecklund and Long 2011;Ecklund and Scheitle 2017;Ecklund et al. , 2016Ecklund et al. , 2019. Notably, within the myriad findings in these studies, one central theme is evident: The relationship between religiosity and scientific occupations is more complex than is commonly believed. Simply, it is not true that being a scientist precludes religiosity. Yet, several scientists evidence extra cognitive burden required to negotiate these dual identities. Additionally, Grant et al. (2016) challenged common misperceptions that medical science displaced religious beliefs about human suffering. Rather, many of the studied health-care workers evidenced a reliance on spirituality in explaining their daily confrontations with death and dying. In summary, the study of occupations challenges common misconceptions about the ways science and religion interact.

Religious Organizations
Within meso-level approaches to studying religiosity, there are two main focus areas: (i) Religious denominations and (ii) religious congregations. These two forms of religious organizations are notably distinct from a third type: Faith-based organizations, or special-purpose groups that involve religion in varying degrees (Ammerman 2016;Fulton 2020). Faith-based organizations are reviewed in a subsequent section. This section attends to the sub-field studying the most overtly religious organizations: Denominations and congregations. There is no singular approach to understanding the boundary markers between religion and secular organizations, as well as how best to define what constitutes a religious organization, such as via mission, staff, or particular programming (e.g., Jeavons 2004;Sider and Unruh 2004). Some scholars treat religious organizations as a distinct field of study (e.g., Fulton 2020; Tracey et al. 2014), while others understand religious organizations to be a subset of nonprofit or voluntary agencies generally (e.g., Bielefeld and Cleveland 2013). Nevertheless, there are some general patterns to the field.

i. Denominations
Alongside the macro-level and micro-level attention to religious tradition, meso-level studies also investigate the role of relatively bounded, belief-based groupings of religious organizations. Though often focused on the same measure, there are important distinctions across level in the conceptualization and meaning of religious tradition. Macro-level approaches tend to view tradition to be a national property distinguishing global cultures, economic systems, and political regulation. Micro-level approaches tend to view tradition to be an individual property distinguishing which groups of people are most likely to hold certain political attitudes, socioeconomic statuses, health and wellbeing outcomes, and so on. In distinction from the other levels of analysis, meso-level studies focus on the role of religious traditions, or denominations, as an organizing body (Warner 1993). This includes attention to the size and scope of religious traditions, by adding to number of adherents a census of congregations belonging to a larger denominational organization.
Meso-level, denominational approaches also attend to the role of leaders, training, size, and other organizational properties in explaining variations in social life. In terms of the distinction between denominations and congregations as the unit of analysis, this is more analytically clean than easily parsed in reality. Analytically, the distinction is best understood by way of example: Attending to all Catholic priests is a denominational approach, since the focus is supra-congregational and religious tradition specific. Alternatively, focusing on all religious leaders (e.g., pastors and youth ministers) within a specific geographic area, net of denomination, is considered a congregational focus, since the denominations are not the primary unit (Snell 2011). Likewise, focusing on organizational properties (e.g., size) within congregations that all share the same denomination is considered congregational, since the sub-denomination organizational properties are the focus. Conversely, studying a set of congregations within a particular area that are compared on key theological, religious-tradition-specific approaches is considered denominational, since the denominations are the central focus.
Admittedly, the lines between these boundaries are porous, and the intention in specifying them is not to reify them. Rather, these are meant to be relatively delineated analytical distinctions to identify a movement across the reviewed studies from a larger organizational unit of analysis-denomination-to a smaller organizational unit-congregations. For a degree of analytical simplicity, denominations are generally understood to be supra-congregations, and congregations to be sub-denominations. After presenting the cleanest descriptions of these approaches, several important exceptions are reviewed.
Denominations are important because they, at least historically, signaled a centralizing authority structure, as well as a level of organization above congregations in which religious leaders could exercise a greater degree of change (Chaves 1993a(Chaves , 1993b, including mobilizing new religious movements within larger religious denominations (e.g., liberation theology within Catholicism: Schwaller 2011). Denominations were also a way to study the market share of particular groups (e.g., Perl and Olson 2000;Sterland et al. 2018), groupings in cultural and political affiliations (Hunter 2001), and variations in subcultural approaches (such as 'strict church theory': e.g., Olson and Perl 2001). Historically, denominations have also been a way to identify major religious changes, such as shifts in identification among the broader population (e.g., Roozen and Nieman 2005) and among younger generations (e.g., Schwadel and Smith 2005).
Thus, several studies focus on collecting censuses of major denominations within a particular geography (typically the U.S.). Despite this rich history of denominational study in the U.S., contemporary trends pose significant challenges to this approach. For example, Roozen and Lummis (2007) identified major denominational splits into smaller supra-congregation affiliations, as well as major disaffiliations of congregations from larger denominational bodies. Moreover, the rise in nondenominational churches is a movement against contemporary relevance of denominations, and their accompanying authority. Yet, complexly, several networks exist to bring together similarly minded sets of nondenominational congregations. While not recognized as denominations, these supra-congregation religious networks provide many of the same functions that denominations did (e.g., conferences, professional training, website branding), though with less formal authority structures. For this reason, several studies focus explicitly on nondenominational congregations and their associated networks (e.g., Thumma 1999;Thumma and Lummis 2007;Thumma 2010aThumma , 2010bSchneider et al. 2010;Smith et al. 2011;Kujawa-Holbrook 2013;Guest 2017;Hall 2016;Brauer 2017;Roberts 2020).
Moreover, these networks have been growing in size and geographic scope. As one example, the Association of Related Churches (ARC) began as a loose organization to build new churches around the country. The movement has now become ARC Global expanding to Canada, Ireland, South Africa, and Australasia. As an additional example, a single congregation emerging in Australia-Hillsong, which began as an evangelical, charismatic megachurch-has now expanded not only through congregations around the world but also as a commercial enterprise through concerts, licensed music, and a network that straddles between a loose denomination and a corporate entity (Marti 2017(Marti , 2018. As a result, the study of congregations is not subsumed within the study of denominations, nor the inverse. Rather, studying religious organizations is best understood as composed of three categories for units of analysis: Denominations, congregations, and special purpose groups (Ammerman 2016), which are also referred to as faith-based organizations.

ii. Congregations
The first step in this approach is to attend to what counts as a congregation. Chaves (2004) defined a congregation as: "a social institution in which individuals who are not religious specialists gather in physical proximity to one another, frequently and at regularly scheduled intervals, for activities and events for explicitly religious content and purpose, and in which there is a continuity over time in the persons who gather, the location of the gathering, and the nature of the activities and events at each gathering" (Chaves 2004, pp. 1-2). With this definition, there is a substantial subfield in the study of religiosity that focuses on congregations (Nieman 2002). Congregational studies expanded in the 1990s and beginning of the 2000s (Ammerman et al. 1998;Ammerman 2016). Then, in the 2010s, critiques arose of the prevalence of congregation studies. For example, Cadge et al. (2011) called for de-centering the study of congregations as crucial to moving the study of religiosity beyond the United States and Western Europe, contexts that are predominantly Christian and congregational.
Nevertheless, studies of congregations remain overwhelmingly focused on Western Europe (Monnot and Stolz 2018) and Northern America, even more particularly on the United States. This is a problem, especially considering American distinctiveness and exceptionalism (Cadge et al. 2011). Yet, due to their sheer volume, attending to extant scholarship on religious congregations necessitates reviewing these U.S.-centric studies. Within the United States, religious organizations constitute the largest segment of the nonprofit sector, and congregations remain the largest subset of religious organizations, with approximately 350,000 to 370,000 (Fulton 2020;Brauer 2017). Congregations receive the greatest share of charitable dollars each year: Over $124 billion in 2018, representing 29 percent of all U.S. charitable giving (Giving USA 2019). Religious congregations also enlist more volunteers than any other sector, and likely, more Americans affiliate with a congregation than any other single type of nonprofit organization (Putnam and Campbell 2012). In summary, congregations are an important organizational entity to study.
Second, another key aspect of organizational studies generally, as well as congregational studies, is to examine their size and scope. This is especially important in congregational studies, since religious congregations are typically not as governmentally regulated as other forms of nonprofit organizations. Several national U.S. studies thus conduct various congregational censusing, such as: While these national congregational studies attend to multiple factors, trends in size and scale in terms of membership and attendance remains one of the most significant (e.g., Sherkat 2014; Bacon et al. 2010;Melton 2010). According to NCS data, the average congregational size has declined over the past two decades, even though the largest congregations have continued to grow. For example, Chaves (2017) reported, "most congregations are small, but most people are in large congregations," as the largest congregations contain increasingly higher percentages of people. Thus, within the subset of congregational approaches, some studies focus in particular on megachurches, defined as congregations with 2000 or more weekend worshippers in one or multiple sites (e.g., Thumma and Lummis 2007;Thumma and Bird 2015). The trend of increasing congregational size has also expanded globally and received scholarly attention in contexts outside of the United States. Indeed, the largest churches in the world are now in countries such as South Korea, Nigeria, and El Salvador (Cheung et al. 2015;Kim 2002;Hong and Myȏng 2011). In attending to congregational size and scope, scholars have also sought to study how the size of congregation matters for levels of religious activity, civic and social engagement, as well as the socioeconomic and education demographics of their attenders .
Third, attending to race and ethnicity is another important area within congregational studies (Edwards 2019). For example, Emerson and Smith (2001) noted the racial and ethnic homogeneity and boundedness of most congregations, and thus the ways congregations can divide by races and ethnicities. Additionally, the 'black church', or predominantly African-American congregations, has received significant attention (e.g., Lincoln and Mamiya 1990;Barnes 2005Barnes , 2009Barnes , 2011Pattillo-McCoy 1998;Nelson 1996). Studies have also focused in particular on Latino churches, or congregations with sizable populations of Latinx Americans (e.g., Hernandez et al. 1994;Marti 2015;Martinez and Tamburello 2018;Matovina 2005;Mulder et al. 2017), and on Asian American congregations (e.g., Alumkal 2008;Jeung 2005). Beyond these studies attending to one or another racial and ethnic group, studies also attend to multiracial or multiethnic congregations: congregations in which less than 80 percent of members share the same racial or ethnic background (Dougherty and Emerson 2018;Edwards 2008aEdwards , 2008bEmerson and Kim 2003;Martinez and Dougherty 2013;Perry 2013a). These studies attend to meso-level issues of religious leadership, power, in-group and out-group dynamics, and how those features shape organizational culture, alongside ethnic identities.
Fourth, there has also been a focus on religious congregations amidst diasporic and immigrant communities (e.g., Warner and Wittner 1998;Ebaugh et al. 2002). For example, studies have focused on congregations as sites of cultural transmission or preservation, forming and sustaining subcultural identities for immigrant groups within a new country, including understanding congregations as sites that aid in preserving language and culture as well as creating social networks within a new country (Warner and Wittner 1998). Additionally, Wind et al. (1998) conceptualized these new religious communities as following principles of institutional isomorphism and adopting a 'de facto congregationalism' with religious leaders, education, and practices that mimic broader U.S. traditions to formalize faith communities into organizations. Alternatively, Cadge (2008) studied American Buddhist temples and found a high degree of diversity in the extent of their congregationalism. All these approaches indicate the ways that congregations, and their associated immigrant and ethnic communities, challenge simple geographic categorization. Instead, Meyer et al. (2011) encouraged studies to pursue greater understanding of the "(1) cross-border interpenetration of religious organizations, beliefs, and practices; (2) variations in the potential for religious beliefs and institutions to be transported; and (3) the use of multiple frames of reference to examine the dispersion of religious cultures and communities" (Meyer et al. 2011, p. 240).
In summary, while overwhelmingly still focused on the United States, studies of congregations in other parts of the world are growing. In the recent past, however, many of these have occurred in additional western contexts, such as in western Europe. Nevertheless, despite the prevalence of publications on these topics, the study of congregations cannot be simplistically labeled as entirely a U.S. or Western European pursuit, nor completely Christian-centric. With careful searching, subsets of congregational studies highlight important elements of diversity study more generally, including particular attention to racial and ethnic dynamics within a nation, and the ways congregations can also be sites of immigration communities and cross-border relations.

Faith-Based Organizations
This section of meso-level organizational approaches attends to faith-based organizations. These are organizations that are not congregations and denominations but still engage religiosity. Notably, faith-based organizations span a wide variety of nonprofit organizational types, from multi-billion-dollar international humanitarian agencies and small local social service agencies to religious student groups on college campuses, arts and humanities organizations, or healthcare providers. The wide diversity of organizational types poses methodological challenges for identifying clear or agreed upon parameters for what constitutes a faith-based organization (e.g., religious language in mission statements, religious affiliations of leaders and boards, or historical coupling with religious denominations). While this inhibits a simple count of the number of faith-based organizations, Fulton (2020) estimated that at least 15 percent of all nonprofit organizations registered in the United States included religious language in their name or mission statement. Plus, Grim and Grim (2016) estimated that faith-based organizations accrued revenues totaling to $378 billion a year. Moreover, Bielefeld and Cleveland (2013) identified at least 600 publications reviewing scholarship on faith-based organizations in recent decades. Thus, faith-based organizations constitute a sizeable subsector, which has received substantial attention from scholars.
Within the broader set of faith-based organizations (and excluding the already covered types above: denominations and congregations), there are at least six additional types. One type is composed of parachurch organizations: Non-church, religion-related, non-profit organizations (e.g., Scheitle 2010; Scheitle and McCarthy 2018; Perry 2013b). Parachurch organizations in the U.S. often engage multiple congregations together around shared purposes. For example, Catholic Youth Organizations organize children and teens from multiple Catholic school parishes in the same geography for the purposes of athletic activities and other extra-curricular clubs. While most often but not necessarily Christian, parachurch organizations are able to operate and expand less through local membership as congregations or denominations but through a shared interest in a particular activity such as education, arts, or social services. A second and related type are large-scale, faith-based umbrella organizations, such as national healthcare systems, accrediting agencies, and professional associations (e.g., Wittberg 2013). Third, another type is composed of inter-faith coalitions that mobilize congregations or religious adherents for civic engagement efforts (e.g., Wood and Fulton 2015;Fulton and Wood 2017). A fourth type is composed of private, grantmaking foundations that explicitly fund grants with religious purposes (e.g., Lindsay and Wuthnow 2010;May and Smilde 2018). Two well-known examples of such foundations are the Lilly Endowment, Inc. and the John Templeton Foundation. Fifth, another type is composed of social service nonprofit organizations that engage faith in some capacity within their service delivery (e.g., Bielefeld and Cleveland 2013;Ebaugh et al. 2006;Cnaan and Boddie 2001). Within social service organizations, Ebaugh et al. (2002) used a Likert-scale to measure the degree of religious influence in decision-making tools, resource preferences, organizational culture, and organizational practices.
A sixth type, related to the fifth, is composed of non-governmental organizations that engage faith within their international aid or development. For example, McCleary and Barro (2006) investigated 1638 U.S.-based, private, voluntary organizations engaged in international relief and development and then parsed the organizations based on secular or religious efforts. Additionally, Clarke and Ware (2015) delineated FBOs as having two identities: "the first is as a development agency seeking to improve the material well-being of the poor, while the second is that of a religious organization whose existence is forged from a faith basis and its understanding of religious tenets  . . . The common factor being that they each are firmly rooted in their religious identity but committed to operating within the development sector" (Clarke and Ware 2015, p. 41). The majority of these studies focus on the U.S. (e.g., Bielefeld and Cleveland 2013). However, as with the studies investigating immigration within religious congregations, attention to faith-based organizations often complicates simple geographic delineation. Additionally, classification of what constitutes faith-based persists. For example, Jeavons (2004) identified the potential for Christian-centrism, even Protestantism, embedded within the terminology of faith-based organization. In response, Sider and Unruh (2004) developed a six-fold typology to parse faith-related organizations into: Faith-permeated, faith-centered, faith-affiliated, faith-background, faith-secular partnership, and secular. This typology is based on an analysis of 12 organizational elements (see Appendix A).
Scholarship on how to delineate religiosity within FBOs, domestically and internationally, demonstrates the difficulty in defining faith-based. Nevertheless, the considerable range of existing studies underscores the need for broader understandings of religiosity and spirituality than macro-level or micro-level studies can provide alone. Attention to meso-level groups and organizations reveals the breadth of these often-overlooked contexts, and their utility for understanding social action. This demonstrates the role that FBOs play in mediating understanding of religiosity and spirituality, not only in local communities and broader networks, but also through individuals. While meso-approach studies have largely focused on religious organizations-particularly congregations, denominations, religious networks, and faith-based nonprofits-there are also increasing calls for identifying religion in "so-called secular venues" such as hospitals, civic organizations, judicial decisions, international diplomacy, or the workplace (Cadge et al. 2011;Cadge and Konieczny 2014). For example, Bender (2011) studied the nature of religion among individuals and in institutions that are secular. Cadge et al. (2011) focused on the role that chaplains play in hospitals, prisons, military, and workplaces in engaging religiosity and spirituality in new contexts. Religiosity in these organizations may be formal or informal, explicit or implicit, and also perhaps fluid and dynamic.

Summary of Meso-Level Approaches
In summary, attending to meso-level aspects of religiosity include studying interpersonal relationship dynamics (e.g., empathy), social networks, moral communities (including atheist and yoga groups), group properties such as size, social trust (enhancing or reducing), in-group and out-group dynamics, occupations, denominations, congregations, and other faith-based organizations. Despite the rich existing information available on mediating structures, meso-level units remain understudied in comparison to the volume of macro and micro approaches. Perhaps this is because large-scale empirical studies have difficulty obtaining meso-level data, or because of other challenges, such as the ability to make meaningful decisions about a single data collection site amidst a myriad of possible organizations and networks, or the need to nuance how religion functions within a diverse set of organizations. In spite of these potential challenges, mediating structures often are key in understanding the intersection of individuals in society. By shaping belief, behavior, belonging, and bonding, as well as channeling individual belief and practice into a broader civil society, groups and organizations facilitate prosocial behaviors such as giving, volunteering, and civic engagement. In terms of the keywords that aid in identifying meso-level approaches, the religious traditions of macro and micro approaches are also present at the organizational level. Additional keywords include: Denomination (then parse by level), congregation (then parse by level), church, mosque, temple, faith communities, religious network, religious organization, faith-based organization, collective prayer, social groups, empathy, and spiritual communities.

Need for Multi-Level Approaches
In a review of 15 years of studying religiosity, Finke (2014) called for both advancement in global studies and increasing attention to multi-level approaches. While recognizing that the distinctions among levels are often porous and mostly constrained by the unit of analysis, Finke also established the need to attend to each level independently, and in intersection. For example, through categorizing the previous decades of research by level, he identified the relative dearth in global meso-level studies. Specifically, Finke stated that little attention had been given to religious social movements and how they are organized, nor to the relationship between religious and social institutions and movements. Notably, he concluded: "To the extent that we ignore this level of analysis, we are handicapping our understanding of religion and the influence of religion on the larger society" (Finke 2014, p. 11).
Thus, one key implication of this review, and Finke's assessment, is the need for more meso-level investigations, alongside continued learning from macro and micro approaches.
Religion and spirituality have a significant role in mediating institutions that shape local communities as well as global networks. Greater attention beyond Northern America to faith communities will facilitate a fuller understanding of the diverse stories of global Christian communities or distinct regional Buddhist or Muslim communities, as well as interactions amidst religiously pluralistic contexts and the development of hybrid religious identities. Plus, meso-level studies demonstrate the power of global religious networks and the complexity of globalization processes. A multi-faceted approach through empirical social scientific research, as well as learning from diverse methods (such as the thick descriptions of ethnography), is necessary to better understand mediating religious institutions and networks. Beyond the overwhelming dependence on Christian-centered and Northern America focused studies in the meso-approach, broader studies of religious organizations large and small would not only improve and broaden knowledge, but it may also help to reshape particular methods that would facilitate greater pluralism in the study of religiosity and spirituality, with attention to fluidity, nuance, and new understandings.

Beyond Western-Centrism
As an initial summary of the above scholarship, two cumulative trends are evident: (1) Decades of study and hundreds of publications compose the broad field of scholars studying religion-a tremendous amount of high-quality studies exist that can inform future investigations, especially for emerging scholars or scholars new to the study of religion; and (2) the field has been plagued by over-attention to Western geographies and religiosities, especially the United States and Western Europe. The latter does not negate the former, yet it seriously conditions it. First, Davie stated nearly a decade ago that religion has and remains "continually present in almost every part of the world" and is "asserting itself in innovative and very visible ways" (Davie 2011, p. 4). Davie continued by stating: "simply deeming religion to be a private matter-the 'traditional' European answer-is no longer an adequate solution" (Davie 2011, p. 9). Second, reviews of the field highlight the parochialism of Western scholars and its impact on the field. For example, Poulson and Campbell (2010) reviewed 490 articles published between 2001 and 2008. They categorized studies based on a combination of Western versus non-Western geographies, and Christian groups versus non-Christian groups. During that time period, studies of Western Christian groups actually increased in proportion, from about 70 percent of the articles to about 80 percent, and the only substantial increase in other combinations was for studies of Western non-Christian groups. Meanwhile, the study of non-Western Christian and non-Christian groups remained stagnantly low (less than 10 percent).
As a response, many scholars in the field call for a change to the Western-centrism of extant approaches. For example, in Ecklund's (2020) presidential address to the Society for the Scientific Study of Religion, a reflection backward on the previous 70 years resulted in a call for a major shift in the coming decades: Diversifying the field, including better attending to the ways that religion is changing globally. Yet, the major unanswered question regards how to globalize the study of religion. Meyer et al. (2011) offered an eloquent description of the need for a 'global imaginary': On one hand, the study of global religion is the study of religion in its global contexts, and involves the study of religious diasporas, the global spread of religious ideas, and the emerging spiritual and moral sensibilities of globalized, multicultural societies (Juergensmeyer 2005). But on the other hand, the global study of religion can affect all dimensions of religious and ideological studies-it involves taking a perceptual stance that is relevant to every aspect of the study of religion, whether the subject matter is local or far away, historical or contemporary, textual or social. The global perceptual stance is one that attempts to see all religious phenomena as part of a global drama, and to understand it through many eyes, from multiple frames of reference. (Meyer et al. 2011, pp. 251-52) Similarly, Yorgason and Dora (2009) highlighted the important ways that the study of religiosity complicates categories and experiences, for example, both by being geographically bounded and by spanning boundaries. Indeed, the study of Islamic religious communities around the world exemplifies this beautiful complexity (e.g., Cherkaoui 2020;Xavier et al. 2018;Rytter 2014;Hamid 2013;Esposito 2010;Spellman 2004). Why, then, does Western-centrism persist? One potential answer is that religiosity has largely been ignored within larger scholarly focus on globalization (Beyer and Beaman 2007). While everything from constructions of gender, understanding of youth, and embedded political ideology have received the benefit of sharp critical attention in contemporary studies, the study of religiosity has remained fairly isolated from this potential for critique amidst broader and more complex global understandings. Perhaps implicit biases inhibit some scholars from considering religiosity to be a valid topic to study, and if so, this limits critiques of Western centrism. To improve global studies, more explicit attention is needed on each level of analysis (e.g., Finke 2014), to assess whether particular levels are lacking within under-studied geographies.
When designing studies collecting multidimensional facets of religiosity and spirituality across different cultures, Traphagan (2005) delineated a set of three important questions when investigating non-Judeo-Christian contexts: (1) How do the people being studied think about religion?; (2) can these subject-identified notions of religion help identify a culturally specific definition of religion that can guide the analysis?; and (3) is the word "religion" an appropriate term to use in this context? Rather than focusing on the formal assumptions in this term, Traphagan emphasized a focus on ritual. This, he suggested, would help researchers move away from the implicit sacred versus profane divide that characterizes much of Northern American and Western European scholarship in this field. To advance beyond this dearth of critical and global attention, the following sections review studies of: (1) Africa, (2) Asia, and (3) Latin America, using United Nations regions (UN 2020; Herzog 2020).

Africa
Within this purposeful sample for studies of Africa since 2010, which attend to religiosity and spirituality, there are a total of 39 publications that are reviewed in this section. Notably, this does not include the macro-level studies reviewed above, which include Africa (or particular African countries) within global, cross-cultural studies. Beyond the summary of those already included above, there is one additional macro-level study included in this purposeful sample of African publications. Dowd (2015) studied the cultural and religious demography of violent Islamist groups within Africa. This approach is aligned with the mainstream trends of other macro-level studies: Attending to national religious oppression or freedom by focusing on religious tradition affiliation.
Within this micro-level research in Africa, two studies focused on the link between religious affiliation and women's livelihoods in Sub-Saharan Africa, which yielded distinct results from the U.S.-heavy micro-level studies above. For example, in a study on religious belonging, religious agency, and women's autonomy in a predominantly Christian area of Mozambique, Agadjanian and Yabiku (2015) found that that women who belong to more liberal Christian denominations, such as Catholicism and mainline Protestantism, tended to have higher autonomy levels. Beck and Gundersen (2016) analyzed the relationship between religious affiliation and earned income in Ghana, considered to be the most religious country in the world. The researchers found that among women, there is a link between denominational affiliation and income, while controlling for education, literacy, region, labor market experiences, and demographic variables. Spiritualists, Pentecostals, and Methodists earned a higher income than Presbyterians, and Traditionalists earned less than all other groups. One theory to explain this variation was the influence of neo-Pentecostal theologies that emphasize wealth accumulation and self-confidence among Spiritualists, Pentecostals, and Methodists. However, self-selection is another possible reason.
Within these meso-level studies in Africa, the focus is often on religious movements, including increases in the number of adherents within existing religious traditions and missional outreach efforts. For example, Dowd and Sarkissian (2017) studied the Catholic Charismatic Renewal movement in Sub-Saharan Africa, specifically within Kenya and Nigeria. Additionally, Offutt (2011) collected ethnographic and interview data to investigate mission trips in South Africa and El Salvador. These rich data reveal how mission trips facilitate cross-border network ties.
In summary, attending to multi-level studies of religiosity in Africa requires searching outside many of the typical journal sources often included in systematic search parameters. This includes attending to journals spanning a range of disciplines (e.g., anthropology, history, religious studies, theology, alongside sociology, political science, psychology, and economics), as well as books, reports, conference papers, and doctoral dissertations. Even with purposefully including such a broad range of sources, Africa remains considerably under-studied, with several African countries not included within a single study. By far the most prevalent approach is macro, with global comparisons among a single or limited set of measures (e.g., religious tradition). Beyond these, micro and meso appear to receive similar levels of attention, but with variations between them by country and source. Cumulatively, these trends mean that limiting searches to one versus another publication outlet is likely to sample studies on particular countries in Africa, as well as constrain the level of analysis. It seems, then, that the only way to engage in meta-analysis of multiple levels, and as representatively as possible within African countries, is to draw wide search parameters. This is not feasible in most studies, especially without funding support, and thus helps to explain how inequalities in attention are replicated across generations, despite concerted calls for change.

Asia
Within this purposeful sample of publications since 2010 attending to religiosity and spirituality within under-studied world regions, studies of Asia are considerably more prevalent than of Africa: 69 Asia publications. For the Asia sample, there are more meso-level publications (36) than macro-level (15) and micro-level (19). In terms of publication type, nine of these are books and another three are book chapters from the following publishers: Oxford University Press (3), University of California Press (2), Stanford University Press, University of Hawai'i Press, Wiley-Blackwell, Springer International Publishing, Routledge, Bloomsbury Academic, and the Institute of Southeast Asian Studies Publishing. Only two of these were not meso-level studies (one was micro and one macro). Of the macro-level publications, five follow a similar approach to the general macro section above: Including Asian countries within a global selection of countries from every world region (Buckley and Mantilla 2013;Luria et al. 2017;Mataic 2018;Van Heuvelen 2014;Wright and Palmer 2018). The remaining 10 macro-level studies, the included countries were: China (2), Taiwan (1), North Korea (2), Sri Lanka (2), India (1), Iran (1), Iraq (1), Lebanon (1), and Pakistan (1). The remaining publication types included one conference paper and 13 articles in these journals: Journal for the Scientific Study of Religion (10) (1), and Personality and Social Psychology Bulletin (1). Combined, the Asian countries studied in all the sampled micro-level publications included: Turkey (7), China (4), Taiwan (3), South Korea (3), India (2), Indonesia (2), Israel (2), Thailand (2) An example of these micro-level Asian studies is Moaddel and Karabenick (2018), who examined religious fundamentalism in eight Muslim-majority countries. They collected representative samples of more than 23,000 adults living in Egypt, Iraq, Jordan, Lebanon, Pakistan, Saudi Arabia, Tunisia, and Turkey. They found that fundamentalist beliefs tend to be stronger in nations where overall religious liberty is low, religion is more homogenous, state power is consolidated, there is strong regulation of religious behavior, and the national context is more isolated from the rest of the world. Individuals that rated high on fundamentalist beliefs tended to have a high confidence in formal religious institutions, belief in religious modernity, belief in conspiracies, xenophobic tendencies, fatalistic beliefs, lower socioeconomic status, relied on less diverse information sources, and held membership in an ethnic majority or the dominant religious sect.
Another micro-level Asian study example is Hu et al. (2017), who found that individuals who identify themselves as Christian in China report significantly higher levels of depression than religious nones or those who identify as Buddhist. The authors concluded that the link between mental health and religious affiliation is contingent on a particular religion's social status. This draws into question whether religious minorities face a psychological burden not found within those who identify with a society's dominant religion(s). Additionally, Gries et al. (2012) conducted a case study of Taiwan that raises questions regarding the wide applicability of zero-sum, forced-choice surveys in measuring religious belief. Gries et al. found that such approaches do not work outside of the Abrahamic context. In countries and contexts where polytheistic beliefs are dominant, positive-sum approaches to measuring multiple religious beliefs are needed. If asked about affiliation, respondents should also be asked to what degree they believe in or practice elements of each of the most common religions in that context. For example, a question such as "Which religion do you believe in?" should be replaced with something along the lines of "How much do you believe in the following religions?" One benefit of this positive-sum approach is the ability to more accurately assess the relationship of different beliefs with each religion. A second benefit is that correlates of belief can be explored more fully. For example, a previous study in Taiwan that took the zero-sum approach gave the impression that there had been a large decline in belief in Buddhism over the past several decades. However, as Gries et al. found, this apparent decline was a result of a change in understanding of the meaning of Buddhism. Further, a shift from zero-sum to positive-sum models may be helpful in Western societies as well. Moreover, Liu (2010) examined the link between risk preference and religiosity in Taiwan. The 'risk-preference' thesis posits that being irreligious is a form of risk-taking behavior. However, this may not be the case outside of the context of Northern America and Western Europe, or even outside Abrahamic faith traditions. In this study, there are essentially two types of religions: 'high-risk' and 'low-risk' religions. Taiwan is dominated by the latter, according to Liu. These religions do not emphasize doctrines or impose strict membership rules, denominational loyalty, or participation in religious activities on a regular basis. Liu found that risk preference is not associated with religious affiliation in societies that are dominated by such low-risk religions.
As a final example of a micro-level Asian study, Sarkissian (2012) investigated the role of religion in civic engagement within Muslim-majority countries. Sarkissian found that religious participation was related to greater political and civic activism. Muslims who attended religious services at least once a week were more likely to be civically engaged than those who attended services less frequently. Notably, neither social trust nor tolerance, variables that are often used to measure social capital in Western contexts, were associated with civic engagement in the studied Muslim-majority countries. The two measures of religiosity were: (1) Religious behavior: Frequency of religious service attendance, and (2) belonging: Membership within a religious organization or house of worship. According to Sarkissian, this two-pronged approach of measuring religiosity raises questions about whether the World Values Survey adequately captures Muslim religious behavior.
As mentioned above, 8 of the 36 meso-level publications are books or book chapters. The remaining 28 publications are journal articles in: Religions (8), Journal for the Scientific Study of Religion (7) Ecklund et al. 2016;Noor 2010;Ismailbekova and Nasritdinov 2012). The majority of these meso-level approaches focused on denominations or congregations, with several attending to other faith-based organizations, more informal religious networks, and mission trips.
In summary, attending to multi-level studies of religiosity in Asia also requires searching beyond the journal sources that are typically included in systematic search parameters. This involves attending to books, reports, conference papers, doctoral dissertations, and journals spanning a range of Asian geographies. Even with purposefully including this broad range of sources, Asia as a whole remains under-studied, and particular Asian countries are not included within a single study. By far the most prevalent approach in this sample of Asian religiosity studies is meso, with attention to congregations, denominations, faith-based organizations, religious networks, and mission trips. Beyond these, micro and macro appear to receive similar levels of attention, but with variations between them by country and source. Cumulatively, these trends further underscore the ways that attending to particular sources or singular-level keywords can bias the collected sample of publications by country. Wide search parameters appear to be necessary to scope all studies of religiosity and spirituality within Asian geographies.

Latin America
Within this purposeful sample of publications since 2010 attending to religiosity and spirituality within under-studied world regions, studies of Latin America are the least prevalent: 30 Latin America publications. For the Latin America sample, there are again more meso-level publications (18) than macro-level (4) and micro-level (8). In terms of publication type, 12 are books from the following publishers: Oxford University Press (3), Cambridge University Press (2), University of California Press (1), University of Chicago Press (1), New York University Press (1), John Wiley and Sons (1), Lexington Books (1), Wipf and Stock Publishers (1), Brill (1). Once again, the majority of the book publications attend to meso-level units. Of the macro-level publications, the only specified country was Brazil (2), as all others attended to the Latin American region as a whole (Casanova 2011;Larrain 2013;Parker 2015;Verona et al. 2010).
Of the micro-level publications, five are articles published in these journals: Journal for the Scientific Study of Religion (3) Within the macro-level studies, one example of an approach is from Verona et al. (2010), who studied the relationship between the religious involvement of mothers and infant mortality rates, within Brazil. The religiosity measures were religious affiliation and frequency of religious service attendance. Affiliation was grouped into these eight categories: Roman Catholicism, Protestantism, Evangelicals, Spiritism, Afro-Brazilian religions, Oriental religions, Judaism, and others. Compared to the Catholic reference group, the results indicated that children born to regularly attending Protestant mothers were more likely to survive. As examples of micro-level studies, Ogland (Ogland et al. 2010;Ogland and Verona 2011) investigated the relationship between religiosity and attitudes within Brazil. Also attending to rising rates of Protestant affiliation within Latin America, the researchers find that while both Brazilian Catholics and Protestants demonstrate opposition to the legalization of abortion, Pentecostals are considerably stronger in evidencing opposition to this.
Within meso-level studies, Audette et al. (2020) compared Catholics with Protestants in Latin America by political participation. Looking at 18 countries in Latin America, those who participated in church activities, beyond simply attending services, were overall significantly more likely to engage in political activities, such as working with others to solve community problems, participating in a protest, and contacting elected officials. However, there were differences along denominational lines. Consistent with U.S. studies focusing on Latinx political participation, Protestant churches in Latin America tended to promote the development of civic skills at higher rates. Yet, unlike the findings of studies in the U.S., Catholics in Latin America who participated in church activities were more likely to translate that religious participation into political action. Additionally, De la Torre and Martín (2016) provided a meta-analysis of religious studies in Latin America.
In summary, attending to multi-level studies of religiosity in Latin America further evidences the need to search beyond typical journal sources in attending to books, reports, conference papers, and journals spanning a range of geographies. Even with purposefully including this broad range of sources, Latin America remains under-studied and appears to be the least studied of the three regions attended to in this review (as compared to Africa and Asia). Moreover, attention within Latin America is uneven and not equally distributed by level. For example, Nicaragua and Argentina received the greatest concentration of attention within these micro-level studies but were not attended to at all within the meso-level studies. Across all levels, Brazil received the greatest proportion of attention in these studies (nearly a quarter of the sample). As with the other two reviewed world regions, meso-level studies composed the greatest proportion of the approaches. This challenges the summary of the field as being most heavily concentrated within macro and micro approaches. Careful attention to geographic scope reveals that this is an accurate summary of U.S. scholarship, and scholarship attending to all world regions simultaneously. Yet, within under-studied world geographies, meso-level approaches are more common than either micro or macro.

Summary of Beyond-Western Scholarship
Cumulatively, attending to the world regions of Africa, Asia, and Latin America revealed that these regions remain understudied, and that inequality exists between and within the regions. Asia appeared to receive considerably more attention, as a whole, than Africa or Latin America. Within the regions, China, India, Brazil, South Africa, and Malawi received considerably more attention than all the other countries within each region. This complicates the description of under-studied and facilitates a focus on within-region inequality in investigative attention. Cumulatively, attending to these world regions indicated the need to extend search parameters beyond the typical journal sources, and to also include books, reports, conference papers, and other scholarly sources. This challenges simplistic systematic search practices and supports the need for purposeful samples (Suri 2011). Indeed, searching within narrow pre-specified sources is most likely to return a heavy dominance of U.S. and Western Europe, as well as bias by level. For example, meso-level studies will be lowest among searches that exclude books. This is likely due to the more complex and intersectional facets of meso-level studies, requiring extended duration in book formats to explain. The results of this analysis confirm the ways that focusing on particular religiosity keywords biases toward certain levels, with the primary cross-level keywords being religious affiliation and tradition.

Discussion
This purposeful meta-analysis of contemporary studies of religiosity and spirituality, and their definitions and operationalizations, reveals the following key findings. First, a plethora of existing scholarship exists that engages high-quality approaches to the study of these topics, which should not be overlooked in future studies. Considerable thought and attention have been paid to how best to define and measure religiosity and spirituality, and the field would advance by better attending to this existing scholarship in considering how to define new approaches and methodological designs. Second, the sheer volume of existing studies presents challenges to synthesis, as does fragmentation across levels, resulting in emerging scholars and interested non-experts not having forthcoming ways to grasp the range and type of extant approaches. To respond to this, the field needs to produce more meta-analyses that attend to definitions across multiple levels, either by including all three levels or by clearly delineating which levels are excluded and why. Third, despite the wealth of existing studies, there are still several improvements that can advance measurement of religiosity (e.g., Finke et al. 2010). Specifically, the field would be improved by greater attention to the strengths and limitations of each level of analysis, as well as by triangulating knowledge across levels. Additionally, the field is still conditioned by heavy attention to Western geographies (U.S. and Western Europe), and tremendous inequality exists in the study of religiosity and spirituality by geography. This includes inequalities in world region attention, and-notably-within-region inequality by country.

Strengths and Limitations
Each level of analysis to studying religiosity and spirituality has a set of strengths and limitations. In terms of strengths, the macro-level approaches facilitate relatively convenient cross-national comparisons on key aspects, such as religious tradition. This approach undergirds attention to disaffiliation, as well as differences and similarities between nations in the relative size of adherent populations within each tradition. Macro-level approaches also appear to appeal the most to scholars who are not specialized within the study of religiosity and spirituality, as is exemplified, for example, by the numerous contributions of the Pew Research Center to widespread public attention in mass and social media outlets to comparative national religious demography. As for limitations, the strengths of one approach are often the limitations of another. Alongside the relative agility and breadth of macro-level approaches is a tradeoff, at least in some studies, with depth and complexity.
For example, the micro-level approaches that include more measures beyond affiliation with a religious tradition raise questions surrounding whether religious tradition affiliation is a mutually exclusive category. The heterogeneity within, and overlaps between, indicate that affiliation may not be compatible with regression model assumptions regarding independence between and homogeneity within variable constructs. Plus, the focus on affiliation could inadvertently perpetuate a Christian and Western centrism that presumes religiosity and spirituality are best measured through highly formalized versions of faith activity, such as congregational participation within particular denomination-affiliated churches, parishes, temples, and mosques. Even without studying congregations explicitly as the unit of analysis, the majority of macro-level approaches focus on religious tradition affiliation, which is predicated on a form of denominationalism that often aggregates participation within congregations into these larger affiliation categories. This approach may be valid within Western contexts in which religiosity is expressed through highly formalized, nonprofit organizational congregations that aggregate relatively cleanly into denominations. However, it may be particularly insensitive to informal forms of spirituality that are more common in non-Western geographies, such as in Africa.
As mentioned, the limitations of one level's approach are usually offset by the strengths of a different level's approach. In this regard, the micro-level approach is stronger in measuring the tremendous depth and rich complexity of religiosity and spirituality, in their many forms. Micro-level approaches evidence that multiple measures of religiosity and spirituality do not necessarily operationalize the same latent constructs. While several scales exist to measure facets of religiosity and spirituality that scale into larger constructs, the volume of scales and measures, as well as the variety of results when these diverse measures and scales are intertwined, indicating that religiosity is not a single underlying factor that can be tapped through a single variable alone. Also, though affiliation with mutually exclusive religious traditions indicates a social phenomenon of value, a plethora of other measures indicate the multifaceted, intricate, diverse, pluralistic, multiplicity that characterizes global religiosity and spirituality. When religiosity and spirituality are studied within local contexts, several limitations emerge for standardizing and efficiently comparing cultures and nations within singular and mutually exclusive categories. The strengths of the greater depth and validity of micro-level approaches thus co-occur with limitations in the amount of time it takes to collect and analyze data, the degree to which locally embedded experts need to be engaged in the research design and analysis, and the lack within the larger field of micro-level studies of orderly, coherent, or lucid conclusions. As a result, findings from micro-level approaches are considerably less compatible for consumption by the public, especially since the results often warrant explanations that require more characters than social media outlets allow, more words than brief mass media articles facilitate, and longer interpretations than the general public may devote attention.
The strengths of meso-level approaches are situated at the intersection between macro and micro approaches. As a supra-micro approach, meso-level studies more readily facilitate comparability across individuals and subcultures, thus undergirding greater coherence in cumulative findings. Alternatively, as a sub-macro approach, meso-level conceptualizations expand the degree of depth and complexity by focusing on the rich interchanges among levels. For example, faith-based organizations that are hosted within one nation and are primarily engaged in efforts within many other nations, such as with international relief non-governmental organizations, are not readily aggregated within one nation for clean between-nation comparisons. In this respect, studying a single international organization itself provides a cross-national study (e.g., King 2019). Yet, meso-level studies are considerably less common than micro-level approaches, within the most typical publication outlets. For example, in the 217 articles reviewed from the Journal for the Scientific Study of Religion since 2010, 104 attended to micro-level units, whereas 66 attended to meso-level units. Similarly, within the 114 JSSR articles that attended to the United States since 2010, 57 studied micro-level units, compared to 39 that investigated meso-level units. Plus, meso-level studies were more common in book publications, which are typically excluded from systematic search parameters.
More generally, the multi-level study of religiosity and spirituality contributes several strengths to the study of social phenomenon. The rich depth, diversity, nonconformity, and incongruity of approaches indicates that religiosity and spirituality receive the complexity that these multi-faceted and sometimes contradictory social forces deserve. Moreover, findings from one level can inform the need for investigations at another level. For example, the macro-level finding regarding the tenacity of religious affiliation to persist as a prevalent group identification in most world regions undergirds the need for more studies at the micro and meso levels to expand understanding of this persistence. Similarly, many meso-level studies indicate the importance of organizational and network ties in widening the reach of national cultures to other nations and regions of the world, offering another lens on understanding the intricate relationship between religion and politics. Moreover, micro-level studies reveal the ways that power dynamics, as part of social life in all times and places, shape the ways that individuals participate in groups, organizations, and nations, as well as how larger social, institutional, and cultural forces simultaneously delimit or control the degrees of freedom for creative and diverse expression.

Future Studies
To advance the multi-leveled study of religiosity and spirituality, the field would benefit from further synthesizing of existing research and scholarship. Of high priority is conducting more meta-analyses of approaches across levels. This will facilitate a novel and fruitful avenue forward that does not perpetuate historical and disciplinary tendencies to only attend to and summarize extant scholarship that focuses on one, but not another, level of analysis. Rather than overlooking or discounting approaches attending to other levels, the field would benefit from cross-level efforts. This does not imply that a single study should, nor could, investigate all levels simultaneously. Rather, the three-leveled framework aids the across-study accumulation of diverse approaches into a broader field of knowledge. Yet, this multi-tiered framework also responds to valid reasons for a lack of coherence or standardization and yields a way to support and value pluralistic methodological approaches. Without more thoroughly attending to scholarship from each level of analysis, the problems with disciplinary silos will likely continue, and entire levels will ongoingly be neglected in synthesizing efforts targeting only one type of study. Advancing beyond these issues will require, for example, attending to multiple publication outlets. It is considerably more efficient to focus only on journal articles, and particular journal outlets, in conducting systematic reviews. Yet, this analysis reveals the flaws with such an approach alone. Future studies need to build toward systematic review that includes a valid set of publication outlets, such as peer-reviewed books as well as journals, and still parses these from grey literature Page et al. 2018;Shamseer et al. 2015). Including a larger set of publication outlets, and in a diverse set of languages, complicates a feasible and manageable set of inclusion and exclusion criteria but doing so is also necessary to avoid potential for methodological and geographical empiricism. This is especially crucial within the study of global nations and cultures as a way to intentionally inhibit implicit biases from one nation or set of cultural values patterning what is considered worthy of study.
One practical implication of this is that the field needs more interdisciplinary approaches. However, one of the characteristics of interdisciplinary research is that the scope parameters are not set or limited by disciplinary standards alone. This is both a challenge and an opportunity. The plan is to develop publication samples that employ systematic parameters, searches, and coding schemes. To promote interdisciplinary systematic reviews, it is necessary to balance the tensions between two important values: (a) Defining the scope wide enough to sufficiently capture relevant publications, and (b) limiting the scope adequately to feasibly and reliably attend to sampled publications. Interdisciplinary publication sampling presents a challenge insofar as any disciplinary consensus on well-regarded journals is not necessarily shared. For example, it is well known within sociology that the top general sociology journals (meaning the journals publish on a range of topics) are the American Journal of Sociology, American Sociological Review, and Social Forces. For example, Smilde and May (2010) include those three general sociology journals as the basis of their general, non-religion-specific, publication sample. There is also general consensus among sociology of religion scholars that the top sub-field, religion-specific journals are: the Journal of the Scientific Study of Religion and the Sociology of Religion. For example, Smilde and May (2010) also include these two journals in their religion-specific publication outlet sample. It is worth noting that a third journal is also sometimes included for applied, practitioner audiences: The Review of Religious Research. Also, many scholars increasingly value a relatively new journal, Religions, due to its international scope and open access format. Open access is key for international accessibility, and thus such sources should be included, even while they are characterized by features of newer publication outlets (e.g., lower impact factors).
When interdisciplinary scholars collaborate, a consensus on valued sources is not as clear. While disciplinary silos can cause problems, one benefit disciplines can provide is a relatively bounded set of publication sources that one needs to read to stay abreast of contemporary trends. Without disciplines defining the boundaries, the flood gates are metaphorically open, in terms of the sheer amount of information that scholars would need to consume in order to gain a sense of the current state of the field. Indeed, this is the sense that many new and emerging scholars have as they begin attempting to acquire domain knowledge, especially if not affiliated with a single discipline and trained through a traditional, disciplinary mentor. One way to respond to this issue is to participate in professional associations that foster interdisciplinary collaborations around a shared topic. The association then aids scholars from different disciplines in gaining access to bounded information that is wide enough to cover a range of approaches, while also remaining narrow enough to provide scholars the ability to stay abreast of the trends within that field. In many ways, associations define the field of network ties and potential collaborations.
For example, an association that fosters multi-disciplinary study of religion is the Society for the Scientific Study of Religion (SSSR). SSSR describes itself in this way: "The Society for the Scientific Study of Religion (SSSR) is an interdisciplinary academic association that stimulates, promotes, and communicates social scientific research about religious institutions and experiences. SSSR was founded at Harvard University in 1949 as the Committee for the Social Scientific Study of Religion. From 1951 to 1956, the organization was called the Committee for the Scientific Study of Religion, and, with membership numbering in the hundreds, the change to the present name was made at the end of 1956. Today, SSSR fosters interdisciplinary dialogue and collaboration among more than 1000 scholars from sociology, religious studies, psychology, political science, economics, international studies, gender studies, and many other fields. The Society's flagship publication, the Journal for the Scientific Study of Religion, is the most cited resource in the field" (SSSR 2020a). While SSSR was formed in the United States and does not explicitly describe itself as international, the current strategic plan conveys a desire to internationalize through the third strategic initiative, which states: "3. Triple the number of underrepresented minority scholars, non-US/Canada scholars and non-sociologists who are members of SSSR" (SSSR 2020b).
With this information about SSSR, and its associated journal JSSR, in mind, one potential way to both expand and limit the scope of publications to review is to focus on journals that publish interdisciplinary scholarship and with international geographies. This approach offers many appeals, yet an application of the approach reveals that the number and scope of such associations and journals is quite wide. For example, another prominent association in the U.S. that fosters the study of religion is the American Academy of Religion (AAR). AAR describes itself in this way: "The AAR is the largest scholarly society dedicated to the academic study of religion, with more than 8000 members around the world. The AAR's mission is to foster excellence in the academic study of religion and enhance the public understanding of religion" (AAR 2020). AAR also has an associated journal, which is called the Journal of the American Academy of Religion (JAAR). JAAR describes itself in this way: "The Journal of the American Academy of Religion is generally considered to be the top academic journal in the field of religious studies. This international quarterly journal publishes top scholarly articles that cover the full range of world religious traditions together with provocative studies of the methodologies by which these traditions are explored" (JAAR 2020). While neither AAR or JAAR are explicitly international in scope, the association intends to build international membership, and the journal publishes articles that are international in scope.
Combined, JSSR and JAAR provide a wide scope of interdisciplinary approaches, with content representing the disciplines of sociology, political science, psychology, religious studies, theology, economics, international studies, gender studies, and likely others. However, to expand the scope to include a publication with a more explicitly international focus, a third source is of value: Religions. Religions describes itself as: "an international, interdisciplinary, peer-reviewed, open access journal on religions and theology" (Religions 2020). This rounds out more of an international focus but limits the sample to those that do not explicitly attend to applied and practitioner audiences. The Religious Research Association (RRA) is known in the field as the primary source of an applied audience and describes itself as having goals to: "Increase understanding of the function of religion in persons and society through application of social scientific and other scholarly methods; promote the circulation, interpretation and use of the findings of religious research among religious bodies and other interested groups; aid in the professional development of religious researchers; and cooperate with other professional societies, groups and individuals interested in the study of religion" (RRA 2020a). While established and primarily focused on the United States, the association also has strategic plans to expand its international scope (RRA 2020b). The associated journal of RRA is the Review of Religious Research (RRR). RRR describes itself in this way: "Provides a forum for the publication of research information on religion and religious organizations in the lives of individuals and in the wider society. Offers a regular channel for the interdisciplinary exchange of information on methods, findings and uses of institutional religious research. Is the only journal to facilitate the sharing and comparing of applied studies between denominational and academic researchers" (RRR 2020).
Cumulatively, this sample of four sources dedicated to the study of religion scopes multiple disciplines, international geographic coverage, academic and applied research. Yet, the field of available sources is not complete. Additional sources that publish religious research include the Sociology of Religion and Journal of Religion, among others. Additionally, books are another important source, especially peer-reviewed university press scholars, as are publications studying the intersection of religiosity with other topics, such as philanthropy and nonprofit studies. Valuable publication outlets to pursue that intersection include Nonprofit and Voluntary Sector Quarterly, Voluntas: International Journal of Voluntary and Nonprofit Organizations, and Nonprofit Management and Leadership. Yet, systematically searching within these publications requires careful construction of meaningful keywords that are designed to populate studies across each of the levels of analysis: Macro, micro, and meso. The current paper provides a step in that direction by conducting a meta-analysis across all three levels and summarizing each alongside their relevant keywords.

Conclusions
In conclusion, this meta-analysis conducted a purposeful search to describe major approaches to studying religiosity and spirituality. The sample consisted of 585 publications, including hundreds drawn from peer-reviewed journals and books. The resulting summaries of major approaches within each level of analysis advance a more cumulative field of knowledge and undergird the ability for future studies to systematically search with a robust set of keywords and search parameters. Based on this comprehensive review, five recommendations are offered to advance the field. First, macro-level studies should include more measures than religious affiliation alone, at minimum including importance of faith and other aggregate measures from the micro-level section. Second, micro-level studies should focus less on the development of entirely new measures and more on the further refinement of existing measures: For example, studying bonding as a feature of religiosity requires more than simple swapping of God with Allah or Yhwh (see Appendix A.2 for a list of existing micro-level measures, as well as Tables A1 and A2 for compilations of existing scales and studies). Third, meso-level studies should advance best practices for measuring faith-based organizations, especially articulating how faith-based determinations can be replicated across domains. Fourth, all studies should consider the multi-leveled dimensions of the investigated topic to assess whether it would be appropriate to attend to more than one level, for example by micro-level studies appreciating the organizational aspects of religiosity and situating individuals within meso-contexts (e.g., asking about how frequently services meet, rather than treating individual-level attendance as within a static organizational frequency, especially considering that rural and less-resourced communities may not have the luxury to offer highly frequent services). Fifth, studies across each level should better attend to geographies outside of the United States and Western Europe, and better narrow the scope of Western-based studies by more clearly explicating the geographic locations in visible meta-data, such as publication titles, abstracts, and keywords.
Despite the saturation of Western-based approaches in available publications, geographies outside of the United States and Western Europe were intentionally scoped in this analysis. Particular attention was paid to Africa, Asia, and Latin America. Attending to these regions revealed that major inequities remain in the geographic scope of religiosity and spirituality studies, including unequal attention between regions (e.g., Asia is studied more than Africa and Latin America) and within regions by country (e.g., China and India receive the greatest attention within Asia; South Africa and Malawi receive the greatest attention within Africa; and Brazil receives the greatest attention within Latin America). Moreover, attention is conditioned by the intersection of geography and level, such that particular countries receive a wealth of attention within particular levels and no attention within others (e.g., searching for Nicaragua and Argentina is likely to return micro-level studies). Thus, future studies need to conduct more meta-analyses, with broader search parameters and careful consideration of geographic scope, and pay greater attention to under-studied world regions, as well as under-studied countries within non-Western regions.

Appendix A. Existing Measures by Level
This methodological appendix includes the wording and response options for examples of existing measures of religiosity and spirituality, investigating each level of analysis: macro, micro, and meso.

Appendix A.1. Macro-Level Measures
Macro-level measures typically ask about affiliation with major world religious traditions. The specific traditions asked about differ by study, with the greatest consistency for the largest traditions.

Micro-Level Measures
Micro-level measures typically ask individuals about belonging, behaving, believing, bonding, salience, and identities. Within belonging, affiliation is measured through similar macro-level religious traditions, with added nuance for within-Christian denominational and non-denominational traditions.
No Identification Additionally, religious disaffiliation (or religious 'nones') are an increasing group, and some studies investigate typical reasons for people disaffiliating with religious traditions.
Parents allowed children to decide whether to continue their affiliation with a religious tradition 2.
Had intellectual disagreements over core issues in religious tradition 3.
Developed doubts about religious tradition 4.
Friends left the religious tradition 5.
Life transitions that disrupted participation (e.g., changed jobs, got divorced, left to attend a university) 6.
Lost a parent or grandparent who was the main influence on the family's religious tradition Resumed religious affiliation 3.

Renewed spiritual interests
Behaving is often measured as frequency of religious service attendance and prayer.
Senior management selection 6.
Other staff selection 7.
Financial support and nonfinancial resources 8.
Organized religious practices of personnel 9.
Religious environment 10. Religious content of program 11. Form of integration of religious content with other program components 12. Expected connection between religious content and desired outcomes Results in 6 types: faith-permeated, faith-centered, faith-affiliated, faith-background, faith-secular partnership, secular. Ebaugh et al. (2006) used a Likert agree-disagree scale to measure the degree of religious influence in decision-making tools, resource preferences, organizational culture, and organizational practices.

•
Service Religiosity: 10 items (Ebaugh et al. , p. 2264 • Sometimes, respondents can choose one of the two options, while in other surveys they have to give a score on a scale from one to ten.

•
In EVS (the European Values Survey), the dichotomous answer format was used in which people could answer: (1) "Most people can be trusted," or (2) "Can't be too careful." • In line with standards in logistic analysis techniques, the "can't be too careful" category is recoded to a score of 0, and therefore, results represent the likelihood of trusting people instead of not trusting people.
• Social Trust: Within-Congregations (Seymour et al. 2014, direct quotes from pp. 133-35): • Our measure of within-congregation trust is a binary measure based on the original item-wording used in the PALS survey and asks respondents if they "have been able to trust completely" members of their congregation during the 12-month period prior to the interview (yes = 1, no = 0).

•
Compared to some other research on the determinants of trust, our dependent measure differs somewhat in both the perceived target and expressed magnitude of trust. • Given our focus on a type of trust that occurs within localized domains, standard questions used to measure generalized trust (e.g., "Generally speaking, would you say that most people can be trusted or that you can't be too careful in dealing with people?") are inappropriate. Instead, the PALS item that we use asks respondents explicitly whether they trust others within their particular congregations.

•
The wording for this item differs from some other ordinal measures used to represent generalized or particularized trust, although it does follow Welzel (2010) and (Wollebaek et al. 2012) in using "trust completely" as a response category to represent the highest rated level of trust.

•
We believe that the wording of the dependent measure is not only acceptable for examining the trust that congregation members have for one another, but may have the advantage of being more focused and less likely to evoke ambiguous, indiscriminate, or unreflective responses from respondents.

•
In future research, it would be useful to compare the specific item wording used in this measure of trust in congregants against other possible variations.
• Religious Diversity (Dingemans and Ingen 2015, direct quotes from p. 746): • Religious diversity equals 1 minus the Herfindahl index of the different religious groups in a country, etc. (i.e., the answer categories are based on those religious groups that exist in a certain country).

•
The nonreligious were included in this measure because they also represent an important and often substantial subgroup in the religious composition of a country. • Our measure correlated highly (r = .81) with the well-known religious fractionalization measure by (Alesina et al. 2003) and by Alesina and Zhuravskaya (2011).
• Religious Leader Relationship (Seymour et al. 2014, direct quotes from p. 135): 1. How close do you feel to the primary religious leader at your congregation? (coded "not at all close" = 0 to "extremely close" = 4; loading = .88) 2.
How often do you talk with the religious leader at your congregation, not including just saying 'hello' after worship services? (coded "never" = 0 to "every day" = 6; loading = .88).
• Religious Network (Seymour et al. 2014, direct quotes from p. 136): • PALS asks respondents to nominate up to four individuals outside of their home that they "feel closest to." • The percentage of a respondent's friendship network identified as a congregation member (ranging from 0 to 100 percent, mean = 26.2 percent) is a measure of relative congregational influence within a friendship network.

•
It should be noted that these nominations are based solely on each respondent's account of these friendships, and we cannot be certain whether these relationships are reciprocated.

•
It seems likely, however, that a congregant's perception of his or her network's composition is sufficient to influence many beliefs and attitudes, including the propensity to affirm trust.
• Religious and Social Otherness (Salvatore et al. 2020, direct quotes from pp. 111-12): • Religious and Social Groups: 3-Islam/Muslim, Immigration/Immigrant, LGBTQ+ Community • Lines of Semiotic Force: 3 (presentation of religious and social groups in relation to mainstream society) 1. Foe vs. Friend: Degree to which social institutions are trusted, have positive connotations, are viewed to be more favorable than were in the past, are reliable, are willing to take care of people's requests 2.
Passivity vs. Engagement: Degree to which people are viewed to be dependent on social institutions, agencies and primary networks; degree to which can cope with an uncertain world; engagement is characterized by the sense of agency, fostered by trust in people and institutions; concerns the meaning of the world as the source of the action directed towards the subject (i.e., passivity) or, in contrast, as the goal of the subject's investment (i.e., engagement) 3.
Demand for Systemic Resources vs. Community Bonds: Concerns conforming to norms and rules due to reliance on those who have power and are part of the majority for functional devices and services needed to address a challenging and uncertain world; versus a need to make life meaningful through vital participation in community bonds that undergirds a sense of agency and control over one's life • Semantic Structures: 2 (similarity vs. difference combined with degree of exposure to otherness) 1. Normality vs. Deviance: otherness as opposed to identity; characterizes a person in said religious or social group as bad, a negative object in opposition to the valorized image of 'people like us' 2.
Domestic vs. Foreign: in the case of Islam, the extent to which was characterized as an issue that needed to be addressed by local, domestic policies versus something lying outside the subjective sphere of action, viewed from an external point of observation, as foreign affairs Combined, Table A1 reports the scales and indices investigated in the reviewed publications for this study.
Combined, Table A2 reports named studies and surveys analyzed in the reviewed publications.