What constitutes a good life? What factors influence it and how it can be promoted? These are important questions appealing to a wide range of people. As the impact of religion/ spirituality on people is robust and generally positive (Emmons and Paloutzian 2003), this study aims to examine the relationship between spirituality and well-being in a sample of participants from Hong Kong, by the use of multiple indicators of spirituality and well-being. In addition, given the enormous influence of religion in the community life, but with only limited attention to the subject (Kloos and Moore 2000), this study would like to explore further the role of sense of community in the mediating mechanisms of religion/spirituality and well-being connection.

Well-Being

There are two major understandings of well-being: hedonic and eudaimonic (Ryan and Deci 2001). While the hedonic stream of subjective well-being (SWB) considered that life’s ideal is to maximize pleasure and happiness (Diener et al. 1999), eudaimonic subjective well-being concerns the notion that true happiness could be found in “the expression of virtues – that is, in doing what is worth doing” (Ryan and Deci 2001, p. 145). While SWB focuses on individual’s feeling towards life and evaluation of life, psychological well-being (PWB) proposed by Ryff (1989) look at people’s positive psychological functioning in daily living, such as having good relations with others or realizing a meaningful life. Recently, there is heated debate whether the differentiation between hedonic and eudaimonic well-being is useful (e.g.,Kashdan et al. 2008), others still argue for including and integrating both perspectives to have a more proper understanding of such a rich concept as well-being (Henderson and Knight 2012; Ryff 1989; Waterman 2008). Despite their differences, the major focus of well-being has been on a personal, subjective realm.

Sense of Community

Well-being is also deeply embedded in the social environment (including family, community and society) and can be nurtured or inhibited by external circumstances (Cowen 1991; Marks 1996; Prilleltensky and Prilleltensky 2006). They have suggested that we have to go beyond the personal and intra-psychic and take into account the personal (such as self-esteem, resilience), interpersonal (having a supportive relationship or family) and collective levels of well-being (enabling sociopolitical condition).

Sarason (1974) proposed the concept of psychological sense of community (PSOC) and argued that for community interventions to be effective, the individual’s feeling of being a part of a larger structure should be the prime focus of researchers. PSOC provides people a feeling of belonging, mutual care, and shared identity (McMillan and Chavis 1986), while empirical studies also showed that PSOC has positive influence on well-being (Davidson and Cotter 1991; Obst and Tham 2009), personal and political trust (Anderson 2010), and civic participation (Chavis and Wandersman 1990).

Religion/ Spirituality (RS)

Conceptualizations of RS may vary according to different disciplinary foci (Smart 1998). In psychology, despite still evolving, the common understanding of RS is that religion and spirituality are related, rather than independent, constructs (Hill and Pargament 2003). In contemporary usage, spirituality concerns the individual and the personal, in contrast with religion which is institutional. Nevertheless, as Pargament et al. (2013) noted, search for the sacred is important for both religion and spirituality, while both are also dynamic, multidimensional and multilevel process. In this study, while we use spirituality in the model building to highlight the core essence and internal aspect of religious life, the authors acknowledge that both religion and spirituality are closely related constructs.

Psychologists often use various dimensional approaches to understanding religion as subjective religiousness (which we termed spirituality). One of which is to understand religion from its innermost experience, including Otto’s (1917) “numinous” experience, James’s (1902/1960) varieties of exotic religious experiences and Pratt’s (1924) “milder type” religious experiences. Despite different foci, as Argyle (2000) notes, religious experience is crucial, and without it, there is no religion at all. Another dimension is concerning religious belief. Park and Paloutzian (2005) and Silberman (2005) both argued that religion, as a belief or meaning making system, is unique because it is not just the assent to certain kinds of verbal propositions, but grounded in powerful emotions towards sacred objects, strong motivation, as well as commitment to action.. Also, there are behavioral/ practice measures of religiousness, including church attendance that is often cited. Hill and Pargament (2003), nevertheless, suggest that the measurement of religion largely depends on what other variables the researcher intends to relate with religion. For example, Hackney and Sanders (2003)’s meta-analysis demonstrate that in the pattern of relationships between religion and mental health, personal devotion (the experiential dimension) has a stronger effect than ideology (belief dimension). In contrast, religious belief (such as fundamentalist belief) is often associated with negative outcome, such as prejudice or dogmatism (Altemeyer and Hunsberger 1992; Saroglou 2002).

Researchers argue that to understand the richness and complexity of religious phenomena, contextually based, multidimensional studies are needed to shed light on the multilayered nature of spirituality and religion (Maton 2001). Dokecki et al. (2001) undertook ten years of action research in a local parish and developed a “head-heart-hand” model of spirituality embracing both human and community development. They argue that spirituality is the pursuit of self-transcendence, going beyond the self, and relating to others, society, the cosmos and finally the Ultimate Other. In relating to the other, and finally the Ultimate Other, spirituality manifests itself in a form of experience of being together (with the divine and the religious community), reflecting on that relationality, and in action or service to strive for ultimate values. Dokecki et al. argue that balancing and integrating the “head-heart-hand” dimensions of spirituality (experience, reflection, and action) is essential for human and community development.

Religion/Spirituality (RS) and Well-Being

Empirical evidence consistently shows the beneficial influences of RS on health and well-being. At the subjective level, a strong relationship between RS and health and SWB was largely supported (Diener et al. 1999; Koenig et al. 2012). Church members were happier than others and less likely to engage in unhealthy or “delinquent” behavior (e.g., drug abuse or alcoholic, pre-marital and extra-marital sex, law breaking offenses). Religiously active people are also healthier, live longer, have less fear of death and have higher levels of satisfaction with life (Argyle 2000; Myers and Diener 1995).

On the interpersonal level, studies support the importance of RS on family life (Holden 2001; Mahoney et al. 2001; Parke 2001), and interpersonal support and relationship (Erikson 1968; Pichon et al. 2007; Rotter 1967). Welch et al. (2004) argued that strong bonding within the church can nurture mutual trust, later extending to beyond the church.

In the collective domain, religion not only generates civic engagement of individual congregants (Wuthnow 1999), but also provides a context that helps build a community. By creating a common symbol and social identity of their own among members, religion can powerfully link the individual to society (Durkheim 200/2001). Empirical research also demonstrates that RS is significantly associated with people’s connectivity in church contexts, including the perception of social support and identification as a church member (Obst and Tham 2009).

In summary, there is strong research evidence indicating the enhancing effects of RS on well-being, at the individual, interpersonal, or collective levels. As Dorr (1984) argued, as spirituality is something that moves a person, the experience of the sacred can be revealed in the personal, interpersonal or even the public sphere, given that a human is a social being. Nevertheless, research on examining the beneficial effect of RS on well-being, using multiple indicators or both construct, is rare.

Sense of Community as a Mediator of Religion-Well-Being Connection

Despite the positive relationship between RS and well-being, the understanding of the underlying mechanisms is still far from clear. For example, Koenig (2005) proposed ten ways that the religion can help improve mental health. They include promoting a positive worldview, helping make sense of difficult situations, giving purpose and meaning, discouraging maladaptive coping, enhancing social support, promoting other-directedness, helping release the need for control, providing and encouraging forgiveness, encouraging thankfulness, and providing hope. Other empirical studies examined the mediating role of meaning in life (Steger and Frazier 2005), forgiveness (Lawler-Row 2010), and optimism and social support (Salsman et al. 2005). Furthermore, social capital and religious social identity were also found to mediate the effect of religion on adolescent moral behaviour (King & Furrow, 2004) and psychological well-being (Blaine and Crocker, 1995; Greenfield & Marks, 2007). These findings demonstrate that RS influences various kinds of well-being through different pathways.

In this study, sense of community is proposed to be a mediator in the relationship between spirituality and well-being for two reasons. On one hand, religion has been found, theoretically and empirically, to help build a common social identity, resulting in a stronger sense of community (Durkheim 200/2001; Obst and Tham 2009). On the other hand, there are also salient relationships between sense of community and various kinds of well-being indicators, such as subjective well-being (Davidson and Cotter 1991), life satisfaction (Prezza and Costantini 1998), personal and political trust (Anderson 2010), and participation in the community (Chavis and Wandersman 1990). In addition, as Chinese culture emphasizes the interdependent view of self (Markus and Kitayama 1991), it is reasonable to expect that the communal relationship can play a critical role in the people’s well-being, even in the religious context. Thus, it is believed that the influence of RS on well-being may be mediated through the sense of community.

A Theoretical Perspective

The present study integrates Dokecki et al. (2001) and Dorr’s (1984) multidimensional perspectives of spirituality and regards spirituality as an integral dimension of a person’s head, heart and hand (i.e., religious beliefs, experience, and practice) within a spiritually-based context. The manifestation of spirituality can be viewed from personal, interpersonal and collective dimensions. The objective of the present paper is to examine the various aspects of spirituality and how they are related to well-being measured by life satisfaction, social trust and sense of community. These variables were employed as outcome variables because both hedonic (measured by life satisfaction) and eudemonic traditions (social trust) can be drawn upon (Ryan and Deci 2001). Furthermore, the social conditions in which a person is embedded (sense of community) will also be taken into consideration. In addition, we address how spirituality is related to well-being and, in particular, the mediating role of sense of community in the relation. The three main hypotheses in this study were:

First, spirituality-related variables would be associated with well-being variables.

Second, sense of community would be associated with well-being variables.

Third, the associations between spirituality and well-being would be, partially or totally, mediated by sense of community.

It should be noted that since various spirituality-related variables, whether belief, experience or practices, are closely related constructs, we do not have particular hypotheses linking specific spirituality-related variable to the outcome variable. Nevertheless, based on Hackney and Sanders’s (2003) meta-analyses finding that personal devotion has a stronger effect than ideology in the relationship between religion and mental health, we expect the experiential or practice dimension of spirituality would be stronger predictors of well-being indicators.

Method

Participants and Procedure

A total of 268 participants (42 % male, mean age = 37.6 years, SD = 12.3) were recruited from 4 Protestant congregations in Hong Kong. As the majority of Christian in Hong Kong belong to the evangelical orientation (Ying 2004), we include both the moderate evangelical and the more fundamentalist evangelical groups.

With the help of the parish workers, the questionnaires were distributed at the Sunday service or some relevant functional groups. Participation in the survey was voluntary, and the purpose, and the nature of the study were stated in covering letters. The consent of the participants was assumed by returned questionnaires and participants anonymously completed a battery of questionnaires.

Instrument

The questionnaires included measures of three dimensions (head-heart-hand) of spirituality (belief, experience and practice) and three well-being indicators (life satisfaction, social trust and sense of community). In addition, respondents’ demographics was also explored. All the scales were translated into Chinese with the meaning of the measurement checked and ensured by a third person with a translation qualification.

Religious Belief

The cognitive dimension of spirituality was measured by the religiosity sub-dimension of social axiom survey (Leung et al. 2002), which refers to the “belief in the existence of supernatural factors and in the impact of religion on people’s lives”(p. 295). The 5 items were presented on a Likert scale ranging from 0 (strongly disagree) to 5 (strongly agree). An example of items is “Religious faith contributes to good mental health.” The Cronbach’s alpha is 0.67 in the present sample.

Religious Experience

The experiential dimension of spirituality was explored by the Daily Spiritual Experience Scale (DSES, Underwood and Teresi 2002), which tapped the ordinary spiritual experience of people. Examples of the scale included: I feel God’s presence”, and “I experience a connection to all life”. Among these 16 items, the first 15 were scored using a modified Likert scale (the response categories are many times a day = 6, every day = 5, most days = 4, some days = 3, once in a while = 2, and never or almost never = 1), as suggested by Underwood and Teresi (Underwood and Teresi 2002). The last item “in general, how close do you feel to God?” has four response categories (not close at all = 1, somewhat close = 2, very close = 3, and as close as possible = 4). The validity of the Chinese version has been supported (Ng et al. 2009), and the Cronbach’s alpha is 0.94 in this sample.

Religious Practices

In terms of the participatory dimension of spirituality, the participants were asked to estimate the time in an average month they spent on 20 major kinds of activities provided by congregations in Hong Kong (Wu and Fok 2006), such as fellowship, choir, visitation or evangelical preaching activities. The total time yielded by summing the response on these 20 activities was examined in the analyses. Due to the violation of normality assumption, a square root transformation was applied to the variable, as suggested by Tabachnick and Fidell (2007).

Satisfaction with Life Scale

(SWLS) (Diener et al. 1985) was used to measure the respondents’ overall subjective evaluation of life satisfaction. The SWLS have been widely applied and validated across nations, including with the Chinese population (Sachs 2003). The Cronbach’s alpha is 0.84 in this sample.

Social Trust

One’s positive relationship with others can be understood by one’s having trusting relationships with others (Ryff 1989). The participants’ level of social trust was assessed by nine items, adapted from Welch et al. (2004). Examples of the items included: “Most of the time people try to be helpful”, and “Generally speaking, most people can be trusted” . The Cronbach’s alpha is 0.73 in this sample.

Sense of Community Index (SCI)

Respondents’ sense of community within their religious community was examined by using an adapted version of Sense of Community Index (SCI, Perkins et al. 1990). The total SCI index has been found to be reliable (Long and Perkins 2003). The Cronbach’s alpha is 0.66 in this sample.

Analysis

The hypotheses were examined by using path analysis. Compared to multiple regression, path analysis examines a more complex model including several dependent variables and mediating variables. Path analysis considers the interrelationship between all variables. Moreover, path analysis can examine direct, indirect, and total effects, and is regarded as full, rather than partial, information analysis (Kline 2005). Path analysis also considers error variance of dependent variables and is commonly used in psychological research (Kline 2005).

As missing values can result in huge data loss, full information maximum likelihood (FIML) was employed to replace missing data. FIML outperforms traditional methods such as listwise and pairwise deletion (Kline 2005), and provides the least biased estimate (Little and Rubin 1989).

Results

The means, standard deviations and correlations for all measures can be found in Table 1.

Table 1 Mean, Standard deviation and correlations among the variables in path model
Full size table

Path Analysis

The proposed mediation path model A (Fig. 1) was evaluated via observed variable path analysis using maximum likelihood parameter estimation (AMOS 22.0). To control the potential confound of age, we also add the variable to the model. Suggested by Meyers et al. (2013), three other fit indexes,Footnote 1 respectively, normed fit index (NFI), the comparative fit index (CFI), and the root mean square error of approximation (RMSEA) were also used, in addition to the chi-square test, to determine the adequacy of fit of the model.

Fig. 1
figure 1

Model A. Structural equation model showing relations among age, spirituality related variables, sense of community and well-being variables

Full size image

The model (A) provided a poor fit of the model to the data (χ 2 = 774, df = 34, p = .005, NFI = .96, CFI = .96, RMSEA = .16). It is worth highlighting here, in Model A, the four direct paths from spirituality-related variables and age to social trust and the direct paths from religious belief to sense of community and life satisfaction were not statistically significant, suggesting that some of the association between these spirituality variables and well-being indicators may not be direct, but be mediated by sense of community.

Thus, a revised model (not shown), with the non-significant paths removed,Footnote 2 was evaluated. Although the chi-square index was significant (χ 2 = 17.80, df = 6, p = .008), the other three fit indices demonstrate an adequate fit of the model to the empirical data of the present study (NFI = .91; CFI = .93, RMSEA = .09). Nevertheless, as suggested by the theoretical ground that positive interpersonal relationship could be a sources of happiness and well-being (Myers 1999; Ryff 1989), a third path model (B, Fig. 2) with the addition of a path from social trust to life satisfaction was evaluated. The chi-square index is not significant (χ 2 = 8.78, df = 5, p = .12, n.s.), in combination with the other fit indices (NFI = .96; CFI = .98, RMSEA = .05), indicating an excellent model fit. The chi-square difference test comparing models B and the revised model indicated that Model B provided a significant improvement in the model fit, χ 2 (1) =10.37, p < .01.

Fig. 2
figure 2

Model B

Full size image

To test for mediation effect, Sobel test was performed to assess the statistical significance of the indirect effect of the mediator in the mediation models. Results also confirmed the mediation effect of sense of community between religious experience and life satisfaction (z = 2.29, p < .05) and social trust (z = 3.14, p < .005), as well as between religious practice and life satisfaction (z = 1.94, p = .053, marginally significant) and social trust (z = 2.38, p < .05). Nevertheless, the mediation effect of sense of community between religious belief and life satisfaction is non-significant (z = 1.56, p = .11, n.s.), and that with social trust is marginal (z = 1.76, p = .08, n.s.).

We further validated the model by performing multi-group analyses, while the same model is analyzed separately to examine if the same structural model is also confirmed in the four congregations investigated in this study. The analysis was conducted in two phases (Meyers et al., 2012). In the first phase, an “unconstrained” or free model, in which the path coefficient will be estimated simultaneously and separately for each group, will be compared with a “constrained” or null model, in which the corresponding path coefficient of each group was supposed to be equal to the other. A comparison of the path loadings of the two models, χ 2 (30, N = 268) =26.79, p > .05, yields non-significant results. That means there is no significant difference in the path coefficient between the four groups and thus we do not need to perform the second phase analysis. The comparison of the covariance of the four exogenous variables, χ 2 (81, n = 268) =291.05, p < .05, yields significant difference. Further examination of the covariance between the groups reveals that only the covariance between age and belief is with group difference, while no group difference was observed for the covariance between spirituality-related variables. For simplicity, only the final model with standardized coefficients was presented in Fig. 2.

Discussion

This study is a rare one that takes the multiple dimensions of spirituality and well-being into account, and the findings can be understood in three major ways.

The Importance of Spirituality on Well-Being

Firstly, our study support that spirituality is associated with well-being in a complex way. Not only is spirituality significantly associated with three well-being indicators, but is the relationship with well-being also diverse, either directly or indirectly or in a combined way, through the mediation of sense of community.

The positive association between spirituality and life satisfaction is not surprising. It is aligned with Hackney and Sanders (2003) meta-analysis that in the pattern of relationships between RS and mental health, personal devotion (the experiential dimension) has a stronger effect than ideology. Nevertheless, the absence of association between spirituality and social trust seems to contradict previous understandings. This study demonstrates an association between spirituality and social trust is fully mediated through the sense of community, suggesting that spirituality still matters, but indirectly through the mediation of sense of a community. This understanding is in line with Newton’s (2001) argument that social trust is more the aggregate of properties of communities, instead of individual members and that religious culture benefits trust. Thus, it may be more beneficial to examine the relationship between social trust and spirituality from the community perspective, rather than from an individual perspective. The positive association between spirituality and sense of community has been illustrated in other empirical work (Obst and Tham 2009). This study further corroborates in Asia, specifically, Hong Kong.

In regards to the association between spirituality and well-being, this study demonstrates that it is largely dependent on the well-being measured. While the relationship between spirituality and sense of community is direct, the relationship between spirituality and social trust is only indirectly mediated through the sense of community. Also, the association between spirituality and life satisfaction is both direct and indirect. This accord with Koenig et al. (2012) who argued that the relationship between spirituality and health is complex and not easy to articulate because not all religious beliefs and practices have a similar influence on health and well-being outcomes. Instead, it depends on the kind of well-being and the dimension of spirituality. All this indicates that the influence of RS on well-being can be richly diverse. Thus, when discussing the relationship between RS and well-being, it is important to state clearly what kind of well-being and which aspect of RS are being discussed. Otherwise, the conclusions can be misleading.

The Mediating Role of Sense of Community in Spirituality Well-Being Connection

Secondly, the mediating effect of sense of community in the spirituality well-being connection is worthy of note. To the authors’ knowledge, this study is the first study examining the mediated effect of a sense of community in the spirituality and well-being connection. Most recent efforts on understanding the mediating link between religion and well-being focus on psychological variables, such as meaning in life (Steger and Frazier 2005), goal (Fiorito and Ryan 2007), forgiveness (Lawler-Row 2010), optimism and social support (Salsman et al. 2005). The present study makes an important contribution to understanding this connection, particularly relevant to the Chinese context. Eckersley (2007) argued that individualism prevalent in western culture brings the considerable cost to health and well-being. RS, with its social and collective focus, appears to have potential to play a significant role in restoring the situation. This study seems to support this in a Chinese context by showing that spirituality may enhance life satisfaction and social trust by building a stronger sense of community. It is particularly relevant to Chinese culture which emphasizes the interdependent view of self, as the feeling of trust and belonging to a community are conducive to people’s well-being. Nevertheless, it is worthy of note that age is negatively associated with sense of community, which looks like a counter intuitive finding. While it is reasonably supposed that people with a longer history in the religious setting (manifested by age) will have a higher level of sense of community, our finding appear to suggest that sense of community is more readily developed in the younger generation than in the older population. Thus, more research, particularly contextually sensitive ones, is needed to explicate the relationship. Furthermore, while the causal effect of spirituality and sense of community and well-being is still to be further corroborated, our findings appear to suggest that spirituality has a unique role in the formation of community and the subsequent effect on well-being in the Chinese context.

The Importance of Social Capital on Well-Being

Thirdly, this study apparently supports the social relation matters in well-being. Path analysis demonstrates that sense of community and social trust combine to affect life satisfaction. On one hand, our findings indicate the critical role of sense of community in life satisfaction and social trust. On the other hand, interpersonal well-being (social trust) was also shown to be significantly associated with well-being at the individual level (life satisfaction). Although both social trust and sense of community were measured in personal level, this findings is in line with the argument that subjective well-being is closely related to one’s interpersonal relationships (Baumeister and Leary 1995) and external circumstance (Cowen 1991). It is also in parallel with the growing literature on the positive effects of social capital on health and well-being (Baum 1999; De Silva et al. 2007; Mckenzie et al. 2002).

In the study of health and well-being, most study have been conducted at the individual level while the effect of the social environment on health and well-being has not been given enough attention (see Marks 1996; Marmot and Wilkinson 1999 for exception). This seems to suggest that well-being is largely individual, with less value attached to concern about the social environment. However, this survey supports the importance of examining contexts, such as social capital, in understanding well-being. Although the definition and conceptualization of social capital is still in debate, many argue that social capital is about the level of network membership, social trust, and sense of community, embedded in a social network (Baum 1999; Lochner et al. 1999; Mckenzie et al. 2002). This study, despite using individual level data, seems to support that the community well-being (feeling a sense of community and social trust) affects subjective well-being (life satisfaction).

Limitation and Future Research

There are several limitations worthy to be highlighted. Firstly, the use of cross-sectional and correlational design has been criticized for its inability to provide causal relationships. While it is reasonably argued that religious belief, experience and practices may affect people’s well-being, one’s life satisfaction and social trust may also influence one’s religious thought, experience and behaviour.Footnote 3 Thus, longitudinal study is better used to clarify the relationship.

Secondly, because of the sample characteristics, more studies are needed to determine if the same mediating relationship can be generalized to non-Chinese populations (which may be more individualistic in nature), non-Hong Kong Chinese (which may not have undergone the colonial history) or other Eastern religions. Future research should also consider multilevel analyses (e.g., congregational type) and including community-based populations who may be spiritual, but not religious/members of congregations.

Thirdly, social trust and sense of community are assessed based on the participants’ self-report and perceptions. Related to this, some believe that sense of community and social trust is an aggregate property and an indicator in the community about the quality of interpersonal or network relationships (Newton 2001), and could be measured at the ecological rather than individual level. Thus, in future research, it would be better to assess the social trust and sense of community by using objective measures and community or social groups as units of analysis, instead of subjective perception of the participants and individual as the unit of analysis.

Last but not least, the understanding and measurement of spirituality is far from easy. Although our survey demonstrates the various ways spirituality may influence well-being, the qualitative difference and the nuance of religious meaning, experience, and behavior in various religious settings cannot be fully captured by survey measurement. For example, there will be a quality difference in the meaning attached to their faith by person who has a social justice sense, as compared to those who hold a fundamentalist belief. While acknowledging that quantitative measurement employed in the present investigation can provide a window to understanding the religious faith, the studied dimensions should be expanded and complemented by qualitative research.

There are some suggestions for further research. Firstly, as well as subjective well-being, the relationship between spirituality and other human strengths (Peterson and Seligman 2004) can be further explored. Secondly, regarding the mechanisms underlying RS and well-being, meaning, goals and optimism are some of the most researched linkages. Examining how multiple pathways (spanning across levels) in the mediating relationship between religion and well-being may be a promising endeavor. Whatever the line of research, an examination of religion and well-being is fascinating because of the numerous relationships that can be revealed and their close relevance to human and community life.