Social vs. Individual Centrality of Religiosity: Research in Religious and Non-Religious Settings in Russia

: Most of the current approaches to measuring religiosity operate with indicators of individual religiosity. One of the central ideas of the current paper is that religiosity is a social phenomenon. The Centrality of Religiosity Scale (CRS) developed by S. Huber is applied to measure the individual component of religiosity. A modification of the CRS (CRS-SOC) has been developed to include the social component of religiosity with two aspects: social connections with lay members of religious communities and with the clergy. The analysis is based on the data of two surveys conducted in Russia: an on-line survey with a general population sample (1768 respondents) and a survey of parishioners of four Christian denominations: Orthodox, Catholic, “traditional” Protestant (Lutheran, Baptist), and the “new” Protestant (Pentecostal) (1192 respondents). The structure of religiosity among parishioners of different Christian denominations is discussed. The results, which revealed that the level of religiosity among Orthodox parishioners is slightly lower, are interpreted using the theory of religious economy.


Introduction
There are so many different approaches to the definition of religion and religiosity (Pollack 2003, pp. 28-56) that some researchers propose to renounce universal definitions of these concepts due to their vagueness and inability to cover all the relevant phenomena (McKinnon 2002). One of the approaches, which allows for the consideration of this complexity (at least to some extent), is the multidimensional approach to religiosity conceptualization and operationalization. This was proposed and actively developed by several scholars in the second half of the 20th century (Glock 1962;Faulkner and Jong 1966;Glock 1959;Fukuyama 1961;Lenski 1961;Fichter 1969;King 1967;King and Hunt 1975;Schoenrade 1991a, 1991b). Probably the most influential of these in the sociology of religion was the approach of Ch. Glock, who proposed studying five dimensions of religiosity: experiential, ritualistic, ideological, intellectual, and consequential (Glock 1962). Later the ritualistic dimension was split into two separate aspects (public and private), and the consequential dimension was removed from the system based on the criticism that it is "logically redundant" (Fichter 1969, p. 172), that it is not a part of religiosity as a phenomenon, but rather follows from it (Stark and Glock 1968). At the same time, in the psychology of religion, G. Allport proposed to pay attention to how central religion is in one's life, whether it is a master motive in one's life (intrinsic religious orientation) or only a means for achieving other goals (extrinsic religious orientation) (Allport and Ross 1967).
One of the most recent developments of the multidimensional approach is the Centrality of Religiosity Scale (Huber and Huber 2012;Huber 2009), which integrates these ideas under a common framework. Huber proposed a model of religiosity, which includes five core dimensions: (1) intellectual dimension-being interested in religious issues; (2) ideological dimension-belief in God or something divine, (3) public practice-participation in religious services; (4) private practice-prayer or meditation; (5) religious experience-oneto-one experience or experience of being at one with all. The model extends the ideas of core religiosity dimensions proposed by Glock (Glock 1962) and also relies on the idea of the centrality of religiosity to personality introduced by Allport (Allport and Ross 1967). The Centrality of Religiosity Scale measures "centrality, importance or salience of religious meanings in personality" (Huber and Huber 2012, p. 711). The scale has been applied in numerous research projects, the largest of which was the "Religion Monitor" survey (Huber 2009).
The Centrality of Religiosity scale considers religiosity as a personal characteristic. At the same time, since the classic works in sociology (Durkheim 1964), it has been emphasized that religion is a social phenomenon, and communication is one of the most important aspects in the persistence of religion (Pace 2016).
The idea of analyzing the social component of religiosity can be found within the multidimensional approach to the measurement of religiosity. Glock and Stark developed a methodology of empirical evaluation of religiosity on Glock's dimensions, which included two additional "relational" indicators: congregational friendships and participation in the activities of religious organizations-the communal involvement. Here Glock and Stark appeal to the ideals of building the Church based on the example of the first Christian communities: "Religious thinkers through the ages have affirmed the ideal that a church should be 'a community of believers', as exemplified by the Christian churches of the first and second centuries. In such an ideal community members are united by strong and intimate personal bonds. Groups in which such close social relations exist constitute what social scientists call a 'primary group' . . . " (Stark and Glock 1968, p. 163). They introduce a continuum with the religious audience at one end and the religious community at the other end. Religious community is characterized by strong interpersonal relations, while the religious audience is just a gathering of people who come to participate in Church services, without any social connections. Nevertheless, Glock and Stark do not include these indicators in the main part of the concept of religiosity, probably because "it has long been recognized that these ideals frequently, and perhaps typically, not fulfilled either by churches or by church members in actual practice" (Stark and Glock 1968, p. 164).
J. Fichter also emphasized the necessity to incorporate the social component (social communion) into the concept of religiosity: "The dimension of social communion (fellowship, associationism, solidarity) means, therefore, that the religious persons we are studying have some kind of affiliation with a religious group, congregation, or church. We need not digress at this point to the consideration of Troeltsch's ecclesia-sect typology, but one of the measurements of religious behavior must certainly represent the extent of church involvement in the sense of social relations with fellow religionists" (Fichter 1969, p. 173).
G. Lenski included the communal involvement (that is the primary relations and fellowship) in the study of religious orientation (Lenski 1961). He differentiated it from associational involvement-that is just church attendance-and found a low correlation between those two indicators.
Although the social component constitutes the core of religiosity, it is rarely included in surveys. For example, in a large study of membership in the Evangelical Church in Germany (EKD), the concept of milieu is used to analyze the situation in the Church (Hermelink and Latzel 2008). However, to a large extent, it can be reduced to individual lifestyle. The issues of communication between the representatives of different milieu in the Church community or the communication of people belonging to the same milieu are not raised.
The social component of religiosity in contemporary quantitative empirical social surveys is sometimes measured by participation in religious community. For instance, it is asked in the World Values Survey in a block of participation in the activities of different voluntary organizations. It can also be analyzed from the point of view of involvement in social networks in the parish or within the congregation. For example, R. Putnam and Ch. Lim, based on the results of the General Social Survey in the United States, have shown that subjective well-being is dependent on the frequency of church attendance, because the latter significantly increases the number of friends in the parish/within the congregation (Lim and Putnam 2010).
Nevertheless, the idea of including the social component of religiosity into analysis is not put into practice nowadays. Researchers usually analyze the influence of individual dimensions of religiosity (most frequently either belief or practice-the frequency of church attendance and prayer) on values, norms and practices. However, the social mechanics of religious values and norms transmission (religious socialization) presupposes communication between the clergy or the core of religious communities and the newcomers-either converts or children born into a particular religious tradition.
This paper aims at returning the social component of religiosity into sociological analysis. A modification of the Centrality of Religiosity Scale, which includes the social dimension, was tested on a sample of parishioners of various Christian denominations and a general population sample in the course of two research projects conducted in Russia in 2017-2019.
According to surveys, most of the population in Russia affiliate with the Russian Orthodox Church. Since 2002, this indicator has not fallen below 50%, and for the last ten years it was within the range 60-80% (Zorkaya 2009;Sinelina 2013;Cooperman et al. 2017;Emelyanov 2016;Chesnokova 2005). One of the most recent surveys, conducted in February 2020, showed that 68% of the Russian population are Orthodox Christians. Other Christian denominations, including Catholic and various Protestant denominations, represent about 1% of the population. The second largest religion is Islam (7%), 16% do not belong to any religion, and 6% are atheists (Attitudes to Religion. Press-Release 2020).
It is well known that the proportion of religious practitioners-active church membersis very low among Orthodox Christians (Sinelina 2013;Cooperman et al. 2017;Emelyanov 2016). This low level of religiosity is usually explained via the close relationship between religious and ethnic or national identity (Cooperman et al. 2017;Karpov et al. 2012), as well as other non-religious factors that replace "true" religiosity (Zorkaya 2009;Filatov and Lunkin 2005;Furman and Kääriäinen 2007). This means that the growth of religious affiliation, observed in Russia since the 1990s, will not be accompanied by the growth of religiosity level. An interesting argument about the limitation for the possibility of growth of the share of Orthodox church-goers "on the supply side" is given by N. Emelyanov (Emelyanov 2016). He argues that the low level of religious practice in Russia among Orthodox Christians will persist for a long period of time because religious supply is strongly restricted by the lack or inaccessibility of priests. The time a priest has for personal contact with parishioners during confession is limited and imposes an upper boundary on the possibility of regular religious practice: "To have the growth of practicing believers reach at least a statistically determinable 3%, it is necessary to increase the body of clergy by 12,620 priests, that is, more than half as much the current number of priests" (Emelyanov 2016, p. 192). Very little is known about the religiosity level of other Christian denominations in Russia. The current paper discusses the religiosity of the adherents of four Christian denominations: Orthodox, Catholic, "traditional" Protestant (Lutheran, Baptist) and the "new" Protestant (Pentecostal).
Most of the discussion about religiosity in Russia concentrates on explaining the gap between high level of belonging and low level of public religious practice (attendance at religious services, confession and communion) among the Orthodox Christians. At the same time, the phenomenon of religiosity is much more complex, at least it also includes private practice, religious experience, and belief. The problem is even more complex, as most of the current approaches to measuring religiosity operate with indicators on the individual level. One of the central ideas of the current paper is that religiosity is a social phenomenon, and this dimension should be accounted for in the research aimed at a holistic approach to the study of religiosity.
The need to distinguish between individual and social components of religiosity is caused by the fact that for certain denominations (e.g., Orthodoxy in Russia) religiosity, measured at the individual level, does not reveal the effects of religion in other areas of life. More religious and less religious people behave in the same way and have the same values (Zabaev et al. 2013;Prutskova 2013). At the same time, the differences in attitudes and behaviors of religious and non-religious Russians can be observed in cases when religiosity is measured not only by the frequency of church attendance or faith in God, but also by belonging to the core or periphery of a parish community. Belonging to a religious community does not only mean a mere increase in religious practice, but it is also about communication and relations with other parish members and the priest. This is when the social dimension allows one to indicate the effects, or consequences of religiosity (Zabaev et al. 2016;Oreshina et al. 2015). It can be assumed that the domination of social components of religiosity (engagement in the community, communication with the clergy, cooperation and mutual help within parish social networks) is typical for many Christian denominations, and other religions as well. These components require adequate measurement tools.

Data Collection Procedures and Sample
The analysis presented in this paper is based on the data of two surveys. Questionnaires of both surveys included the same block of questions according to modified version of the Centrality of Religiosity Scale (CRS-SOC).
The first was an online survey "The Paradox of Interrelation between Religion and Family in Modern Russia" (1768 respondents) conducted on 27-29 May 2019. The fieldwork was carried out by the Tiburon Research company in Russia. The respondents were randomly selected from an online data base of more than 500,000 people. It was a quota sample (controlled for gender, age, and geography), representative of Russians who have access to the internet.
The second one was a survey of parishioners of four Christian denominations in Russia: Orthodox, Catholic, "traditional" Protestant (Lutheran, Baptist) and the "new" Protestant (Pentecostal). The research project, "Religion and patterns of social and economic organization. Elective affinity between religion and economy in Christian denominations in Switzerland and Russia" was carried out in 2017-2018. The surveys were conducted in various regions of Russia after the liturgy/mass/Sunday services.
The search for and selection of respondents was done by means of contacts with representatives of religious organizations. Since highly religious respondents are rarely included in the samples of mass surveys, for studying the members of this group it was necessary to collect the data at survey points where such respondents would be present in sufficient numbers. Conducting the survey as people were leaving religious services fully met these objectives.
The parishioners were surveyed as they were leaving the liturgy, mass or Sunday worship. By prior agreement with the parish priors or bishops at the end of the liturgy, mass or Sunday worship, the clergy informed worshipers about the survey and asked them to complete the questionnaires. Announcements about the survey were also posted on the doors at the entrance to the church and, if possible, on information boards. The questionnaires were handed out to parishioners at the exit of the church by three to five interviewers. The questionnaire was filled in by the respondent and, if feasible, was submitted right away. If the respondents were in a hurry to leave, they were asked to fill in the form at home and to return it during the following week.
This method of data collection made it possible to obtain a sample of medium and highly religious respondents (aged 16 years or older). In total, there were 18 parish surveys, and an additional online survey of parishioners from religious groups that were hard to reach (24 parishioners, mostly Catholic, were interviewed using the online version of the questionnaire). The total sample size for the analysis was 1192 respondents.

Conceptualization
In our study, religiosity is measured via the Centrality of Religiosity Scale (CRS) developed by S. Huber (Huber and Huber 2012;Huber 2009). According to Huber and Huber, five core dimensions should be distinguished (Huber and Huber 2012, pp. 714-15): • "From a sociological perspective, the intellectual dimension refers to the social expectation that religious people have some knowledge of religion, and that they can explain their views on transcendence, religion and religiosity. In the personal religious construct system this dimension is represented as themes of interest, hermeneutical skills, styles of thought and interpretation, and as bodies of knowledge . . .

•
The dimension of ideology refers to the social expectation that religious individuals have beliefs regarding the existence and the essence of a transcendent reality and the relation between the transcendence and human. In the personal religious construct system this dimension is represented as beliefs, unquestioned convictions and patterns of plausibility . . .

•
The dimension of public practice refers to the social expectation that religious individuals belong to religious communities which is manifested in the public participation in religious rituals and in communal activities. In the personal religious construct system this dimension is represented as patterns of action and as a sense of belonging with respect to a certain social body as well as to a certain ritualized imagination of the transcendence . . .

•
The dimension of private practice refers to the social expectation that religious individuals devote themselves to the transcendence in individualized activities and rituals in private space.
In the personal religious construct system this dimension is represented as patterns of action and a personal style of devotion to the transcendence . . .

•
The dimension of religious experience refers to the social expectation that religious individuals have 'some kind of direct contact to an ultimate reality' (Stark and Glock 1968, p. 126) which affects them emotionally. In the personal religious construct system this dimension is represented as patterns of religious perceptions and as a body of religious experiences and feelings." In the current study, some modifications were introduced into the traditional version of the Centrality of Religiosity Scale. Based on the distinction of individual and social components of religiosity, a modification of the CRS (CRS-SOC) was developed to include the social component of religiosity.
In religion, individual and social components are closely connected. On the one hand, the social components form the religious profile of an individual. On the other hand, individual and social components are simultaneously present in basic religious activities. Prayer, for example, is an individual action but, at the same time, prayer is performed in the language of the previous generations, making it a social phenomenon (Mauss et al. 2008).
However, the individual and social component can be theoretically separated by introducing a particular conceptual border between them. The treatment of this problem proposed in the current paper is similar to the approach used in Social Network Analysis (Wasserman and Faust 1994), which focuses on the relations between actors, and not on the attributes of actors. Similarly, in our study, the individual component of religiosity is the attributes of individuals (believing in God, religious practices, thinking about religion, and religious experience), while social relations between different actors (individuals, communities, churches, etc.), based on religiosity, constitute the social component of religiosity.
There are different versions of the Centrality of Religiosity Scale. The shortest version (CRS-5 1 ) includes five indicators (one item for each dimension). A longer CRS-10 version includes ten indicators with two items measuring each dimension, and the CRS-15 has three items per dimension. The interreligious versions of the scale (CRSi-7, CRSi-14 and CRSi-20, respectively) include additional items to account not only for the dialogical pattern of spirituality, but also for the participative pattern. These are the frequency and importance of meditation, the religious experience of being at one with all, the feeling of being touched by a divine power, or trying to connect to the divine spontaneously when inspired by daily situations (Huber and Huber 2012, p. 717).
In the longest versions of the scale (CRS-15 and CRSi-20), the dimension of public practice includes not only the frequency and importance of participation in religious services, but also the importance of being connected to a religious community. This item directly corresponds to what I propose to consider a separate social component, or dimension of religiosity. When shorter versions of the scale are used in research, this component is completely lost, but the most important reason for splitting the public practice dimension with extraction of the social component is that a social connection is not practice per se. It can be created by, or manifested in practices of communication, mutual help and social support, exchange of information or other resources, but theoretically, it is different from practice (repeated activity). This repeated activity can be performed individually, even if it is public (visible to others). Social connections go beyond the individual level, as they include relations, interdependence, and a sense of belonging or attachment to a social body (community) or to other individuals. These connections can either be with "religious professionals", who conduct religious services and provide spiritual guidance (the clergy), or with religious communities (the laity).
In various denominations, individual and social components of religiosity are present and accentuated in different proportions. For example, M. Weber argued that the doctrine of predestination in Calvinism resulted in the spiritual isolation and loneliness of a solitary individual. Believers "were directed to pursue their life's journey in solitude" (Weber 2011, p. 119), and the Puritan literature often warned against friendship and trust in others. Accordingly, independence and self-reliance is an important characteristic of many "traditional" Protestant denominations. At the same time, other Christian denominations put great emphasis on building strong church communities-the Pentecostals are a vivid example (Martin 2002).

Operationalization
The items for each dimension are presented in Table 1. Both English and Russian versions are provided.
There was a slight difference in the wording of the Public practice (attendance at religious services) item between the samples. The Russian phrase "религиoзные службы" was used as a translation of "religious services" in the questionnaire, presented to the general population sample. This translation is universal and suitable for different religions, although for church-going Christians, it might sound too formal (and even alien to some extent). In the questionnaire presented to Christian parishioner sample, the word "бoгoслужение" was used instead. This is closer to the respondents' language.

Data Processing Issues
The denomination of a respondent was determined on the basis of self-identification, i.e., answers to the corresponding question in the questionnaire, and not on the basis of survey locations (religious service of a particular denomination). This decision was not obvious; it was made after the survey. Initially, we assumed that we would find Pentecostals in the Pentecostal church, Catholics in the Catholic church, etc., but after the survey, it turned out that this was not always the case. For example, several Muslims and Buddhists were present at Catholic services, Baptists were at Pentecostal services and vice versa, there were also those who did not belong to a particular denomination (e.g., just "Christians", atheists, or non-affiliated people who came to the church as tourists, to accompany a more religious partner or spouse, etc.). Such questionnaires were filtered out from the database. An additional difficulty was that the sample included a small number of respondents with double affiliation; in all cases one of them was Orthodoxy ("Orthodox Catholic" or "Orthodox Protestant"). Since such cases were not enough for a separate statistical analysis of this group, a decision was made to classify such respondents not to Orthodoxy, but to the second denomination indicated.

Ideological dimension
To what extent do you believe that God or something divine exists?
Three versions of the scale were constructed and compared: (1) the traditional CRS-5, (2) the scale with only one item measuring the social dimension (social connections with the clergy)-the CRS-SOC6, (3) the scale with two items measuring the social dimension-the CRS-SOC7. The scale values were computed via averaging the corresponding item values.

Results
Scale reliabilities (Table 2) are quite different in each of the samples in our study. Cronbach's Alphas are highest in the online survey with the general population sample (0.85 for both the CRS-5 and the CRS-SOC6, 0.79 for the CRS-SOC7) which allows us to conclude that the reliability of both scales is good, with the CRS-SOC6 performing a little better than the CRS-SOC7. Both items of the social dimension have a highly skewed distribution in the general population sample (Table 3). Many respondents gave negative answers to the question of the sense of belonging to a particular church community (90% said "not at all"). The distribution of answers to the question on social connections with the clergy is a little less skewed (73% said "not at all"). The scale with six items, measuring the social dimension with only one indicator (the social connections with the clergy), was just as reliable as the traditional CRS-5 in the general population sample.
If the parish sample is considered, the reliability of all scales decreases, compared to the general population sample, although it stays within the acceptable range (0.60 for the CRS-5, 0.63 for the CRS-SOC6, and 0.68 for the CRS-SOC7). The effect of decreasing reliability in parish samples can be due to the fact that, in these samples, there are almost no respondents with low religiosity levels (Table 3), measured with the traditional CRS-5 indicators, so the correlation between single items is calculated based not on the whole range of possible values, but only on the positive poles of the variables. This leads to the situation that even slight incongruences in the salience of dimensions play a much greater role in the parish sample than in the general population sample.
The addition of the social component strongly increases the scale reliability in terms of internal consistency in the samples of parishioners of all denominations, especially the Orthodox and the Pentecostal. Here the CRS-SOC7, which includes two items measuring the social dimension, performs best.
Higher reliability of the CRS-SOC7 in the parish sample can also be explained if the linear distributions of each item are examined (Tables 3 and 4). The situation is perfectly reversed as compared to the general population sample. In the parish sample, all items traditionally used in the CRS-5 have a highly skewed distribution. Most of the respondents are characterized by high values on the scale. Only the social dimension items have a significant variance. These are the only items, which differentiate the respondents well.  The correlations between separate items are much higher in the general population sample (Table 5) than in the parish sample (Table 6).
The inclusion of the social dimension allowed for the construction of a scale, which is as internally consistent as the traditional CRS-5 in the general population sample but performs much better in a specific sample of churchgoers.
In the next section, the religiosity of the parishioners of the four Christian denominations is described according to particular dimensions of the individual and social component of religiosity and compared to religiosity of "average Russians" (Table 7).
Based on the previous studies and arguments discussed in the introduction, it could have been expected that the religiosity of Orthodox Christians would be the lowest of the denominations in our study. If we compare Christian denominations in the parish sample, based on the total centrality of religiosity scores, and the raw items, Protestants of various denominations turn out to be slightly more religious than Orthodox Christians and Catholics. At the same time, the difference is not as huge as might have been expected.
The structure of religiosity differs slightly among Christian denominations participating in the research, based on ANOVA and Tamhane criterion for multiple comparisons (Tables A1 and A2 in Appendix A). This means that a particular religiosity dimension can be more or less pronounced within a particular faith. Among the dimensions, which constitute the core of the Centrality of Religiosity scale, only the public practice dimension allows differences between Catholics and Orthodox to be seen. Catholics attend church services more often.
Protestantism is expected to be more individualistic than Catholicism and Orthodoxy. "Traditional" Protestants are actually slightly higher on the intellectual dimension than parishioners of all other Christian denominations. They think about religious issues more often than others. On the other core dimensions, they differ only from Catholics and Orthodox, but not from Pentecostals.
The social component of religiosity and, particularly, the parish community component (the relations with lay parishioners), revealed the highest absolute differences between the four Christian denominations in our study. Contrary to expectations of greater individualism of Protestants, they are higher on this dimension than all other denominations. This is probably a result of the specific Russian situation with their ethnic identity. Lutherans in Russia often have German origin. Belonging to a parish community is a part of belonging to an ethnic minority group for many of them. Participation in non-liturgical activities of the parish allows for preserving national holidays and traditions. Our study shows that the social component of religiosity is also characteristic of Pentecostals. The Pentecostal religious service is much less formal than the Catholic or Orthodox service, and often presupposes the active participation of lay church members.
Theoretical expectations of the greatly lower religiosity of Orthodox parishioners are not supported by empirical data, although moderate differences with representatives of other faiths are nevertheless observed. At the same time, the religiosity of all four Christian denominations in our parish sample differs greatly from religiosity level in the general population sample. This means that not only the minority denominations (Catholic, "traditional" Protestant, and Pentecostal) constitute a specific group, but the Orthodox churchgoers do as well. The paradox we find in our study is that, within the majority religious group in Russia, there seems to exist a minority group very similar in terms of intensity of religiosity to other minority groups. The religiosity of the Orthodox Christians is very different, depending on the point from which it is observed. From the point of view of social scientists, based on the data of general population surveys, Orthodox religiosity is quite weak. At the same time, churches are overcrowded with quite highly religious parishioners.

Discussion
Based on previous studies and interpretations proposed by sociologists, it could be easily explained why the religiosity of Orthodox parishioners is slightly lower than among parishioners of other Christian denominations. Firstly, this may be the result of "ethnodoxy"-a mixture of religious and ethnic identities characteristic of Russians: "Ethnodoxy is a belief system that rigidly links a group's ethnic identity to its dominant religion and consequently tends to view other religions as potentially or actually harmful to the group's unity and well-being and, therefore, seeks protected and privileged status for the group's dominant faith" (Karpov et al. 2012, p. 644). Many sociologists note that, among the Orthodox, there is a large proportion of those who identify with the Orthodox Church solely, according to the principle of either national or ethnic identity ("I am Russian, therefore I am Orthodox.") Moreover, such a mixture of denominational and national identity may weaken or intensify due to the influence of different factors. For example, Kublitskaya notes that "confessional identification in the region increases with increasing conflict in interethnic interaction" (Kublitskaya 2016, p. 116). However, in addition to this reason, other arguments can be proposed, based on a different logic of justification, and which involve other theoretical resources. For example, G. Davie introduces the concept of "vicarious religion". Vicarious religion involves "delegating" the obligation to believe, regularly perform religious rituals, and lead the church lifestyle to "an active minority but on behalf of a much larger number, who (implicitly at least) not only understand, but, quite clearly, approve of what minority is doing" (Davie 2006, p. 22). This effect should also be characteristic of a dominant denomination rather than religious minorities. Nevertheless, the above arguments are valid only when explaining the low level of religiosity of nonchurch-going Orthodox Christians who do not confess or receive Communion, and do not fulfill other requirements of the church lifestyle. For these reasons, such people were not supposed to be included in our sample as we tried to construct the sample of active churchgoers. At the same time, when it comes to "churched" Orthodox Christians (Sinelina 2013;Chesnokova 2005), whose lifestyle includes regular religious practices, other theoretical grounds are needed to explain the differences in their religiosity level from the parishioners of other Christian denominations. The rational choice theory can provide such grounds. According to some theorists of religious economy, members of strict or conservative religious organizations that are in tense relations with their social environment are generally more religious (Iannaccone 1994;Stark and Finke 2000). The theory of rational choice is based on the premise that people compare the costs and rewards of participating in the life of a religious organization and choose the degree of religious involvement that minimizes costs and maximizes rewards. Membership in religious groups with higher tension is "more exclusive, extensive, and expensive" (Stark and Finke 2000, p. 145), and at the same time it can be more rewarding due to this exclusiveness. Small and demanding denominations are able to overcome the free-rider problem, that is benefiting from religious membership without contributing (Stoll and Petersen 2008). Thus, the theory of religious economy provides more reasonable arguments that allow us to interpret the results of our study, which showed that the level of religiosity among Orthodox parishioners is slightly lower than the level of religiosity of parishioners of other Christian denominations.
The theory of religious economy emphasized the importance of social relations between a religious community's members and the population outside the church fence. This is another topic yet to be explored in research on the social dimension of religiosity. If the religious denomination represents a minority, and the relations with the outside social world are tense or constrained, it leads to an increase in intensity of in-group social connections, even in denominations which have a quite individualistic doctrine. This was the case with the "traditional" Protestants in our research.
Individuals are not autonomous. They are involved in different types of social relations. The social relations formed on the basis of religion constitute the social component of religiosity. These include relations with lay parishioners or religious community members and relations with the clergy. These relations can differ: acquaintance, interaction, mutual assistance, participating in Church voluntary activities apart from religious services, etc.
In various denominations, individual and social components of religion are present and accentuated in different proportions. However, at the same time, it is important to notice that collective rituals, or the public practice dimension of religiosity (Huber 2009;Huber and Huber 2012), is not the social component in itself. Church attendance does not automatically produce social relations, and the phenomenon of "sacred individualism" (the attitude that one comes to church to communicate with God and not with other people) in contemporary Russian Orthodox Christianity (Zabaev 2011) could serve as an example. A holistic approach to the operationalization of religiosity should include a measure of social relations formed on the basis of religion.
Previous research has shown that the indicator of belonging to a parish community has a significantly higher explanatory power than individual religiosity dimensions in the research of religiosity effects on values, attitudes and behavior. This research would be more sophisticated if it were equipped with a reliable method of religiosity measurement, which included the social dimension. It can be recommended to use the CRS-SOC6 version of the scale in general population surveys and CRS-SOC7 in surveys of churchgoers.

Funding:
The data was collected in the course of two research projects. The project "The Paradox of Interrelation between Religion and Family in Modern Russia" was supported by the Russian Science Foundation in a form of a grant (project № 18-78-10089). The grant was given to Saint Tikhon's Orthodox University. The research project "Religion and Patterns of Social and Economic Organization. Elective Affinity between Religion and Economy in Christian Denominations in Switzerland and Russia" was supported by the grant of the Russian Foundation for Basic Research (RFBR) No. 16-23-41006.

Conflicts of Interest:
The author declares no conflict of interest. The funders had no role in the design of the study; in the collection, analyses, or interpretation of data; in the writing of the manuscript, or in the decision to publish the results.