Delight or distraction: an exploratory analysis of Sabbath-keeping internalization.
Sabbath keeping--substantial time set aside each week for rest, worship, and renewal--is a core theme in the Old Testament and the Gospels. The commandment to "remember the Sabbath day to keep it holy" and to "not do any work" frames the Sabbath as a remembrance of God as Creator (Gen. 2:3; Exo. 20:8-11) and Deliverer (Deut. 5:12-14). Rest, moreover, is for the community--including servants and livestock-and even for the land (Lev. 25:4). Even before the Ten Commandments were given on Mt. Sinai, the provision of manna on only six days of the week (Exo. 16) required a daily act of trust--for five days, the Israelites had to trust that the manna would be provided for the next day, on the sixth, they had to trust that the extra portion would not spoil. Thus, the intent of the Sabbath commandment is not only rest from daily toil, but rest that strengthens individual and community dependence on God as Creator, Deliverer, and Provider--Sabbath as freedom and a delight (Luke 13:10-17; Is. 58:13-14).
However, another experience of Sabbath keeping exists-the experience of coercion and distraction. Amos 8:4-6 (New International Version) describes this mindset: "Hear this, you who trample the needy and do away with the poor of the land, saying, 'When will the New Moon be over that we may sell grain, and the Sabbath be ended that we may market wheat?'"
Here we find apparent Sabbath keepers whose attentions are focused elsewhere--imagining themselves out from under the perceived thumb of the Sabbath, so that they might return to profiteering against the powerless. Amos relays God's judgment: an object lesson whereby the perceived constraints of the Sabbath are traded for very real suffering. This ancient prophetic accusation parallels some current psychological studies of well-being in demonstrating how decreasing materialism increases well-being while increasing materialism does the opposite (Kasser et al., 2014).
The experience of Sabbath as either delight or distraction--based on the mindset of the particular Sabbath keeper--reflects a more general pattern in religion. Religious beliefs and behaviors have been associated with positive social outcomes (Pichon, Boccato, & Saroglou, 2007; Xygalatas et al., 2013) and psychological health (Maltby & Day, 2003) as well as higher levels of subjective well-being (Hackney & Sanders, 2003; Myers, 2000). However, Mochon, Norton, and Ariely (2011) found that while believers with strong commitments to religion reported the highest levels of well-being on average, the lowest levels of well-being were found among believers with moderately low commitments--not among the non-religious. Moreover, the relationship between religiosity and subjective well-being is attenuated in societies with better living (that is, material) conditions (Diener, Tay, & Myers, 2011). These patterns suggest both that it is the psychological experience of religion that matters for well-being-happy and unhappy believers share the same pews-not the mere exposure to religion per se, and that the internalization of practices that bring one either closer to God or closer to greed are in direct conflict (Matt. 6:24).
This is also consistent with four decades of literature chronicling the continued growth of conservative or strict churches with costly religious practices (Iannacone, 1994; Kelley, 1972; Thomas & Olson, 2010). (1) High-cost religious rituals and practices have been identified as purported means of signaling cooperativeness and building prosociality in religious groups (Sosis & Bressler, 2003; Xygalatas et al., 2013), but also as means of simultaneously meeting human needs for belonging and distinctiveness (Hornsey & Jetten, 2004). It appears that the consistent repetition and internalization of distinctive, meaningful, and effortful religious practices cements membership (Thomas & Olson, 2010) and perhaps increases well-being--at least for those who internalize those practices and remain in the religious community (Iannacone, 1994). Indeed, believers raised and remaining in religions with multiple high-cost practices and distinctive subcultures report better health than those in lower-cost religions as well as those who switch out of the high-cost groups (Scheitle & Adamczyk, 2010).
Previous work examining the relationship between internalization of religious motivations and well-being (Neyrinck, Vansteenkiste, Lens, Duriez, & Hutsebart, 2006; Ryan, Rigby, & King, 1993; Zeldman, 2006) demonstrates that when religious practices are seen as autonomously regulated, congruent with personal goals, and inseparable from the self, believers report higher levels of overall well-being. When religious practices are controlled by guilt, social pressure, fear, or shame, believers generally report lower levels of wellbeing (Dudley, 1978; Mochon et al., 2011; Neyrinck et al., 2006; Ryan et al., 1993) because of increased stress (Weinstein & Ryan, 2011). However, these previous studies of internalization have not typically focused on distinctive, high-cost practices held by a few strict churches--practices such as Sabbath keeping--but rather on widespread low-cost practices (Ryan et al., 1993; Zeldman, 2006) or self-generated low-cost practices (Neyrinck et al., 2006). Thus, we do not know if internalization of the costly practices that are most meaningful to a religious community has effects on well-being apart from the internalization of low-cost practices.
Sabbath keeping is, of course, a high-cost practice especially when Sabbath keepers disconnect from the prevailing economic system for a full day each week (Bass, 1997; Brueggemann, 2014, Diddams, Surdyk, & Daniels, 2004; Superville, Pargament, & Lee, 2013; Waller, 2009). Not only does devoting an entire day to separation from the secular incur an economic cost, but there are likely also (at least for some believers) attendant psychological costs of suppressing thoughts related to secular concerns. The high cost of Sabbath keeping may be one reason that it is a relatively rare Christian practice (Bass, 1997; Brueggemann, 2014; Bull & Lockhart, 2007; Diddams et al., 2004; Waller, 2009); a full day of Sabbath rest is typically only found among orthodox Jewish communities or a few explicitly Sabbatarian Protestant denominations. At the same time, interest in Sabbath keeping as an intentional Christian practice is growing (Bass, 1997; Diddams et al., 2004; Superville et al., 2013)--although Christians who desire a Sabbath rest often lack the support of a Sabbath-keeping community.
In the first major empirical study of the relationship between Sabbath keeping and well-being, Superville and his colleagues (2013) found small but significant total and indirect effects of Sabbath keeping (via religious coping, religious support, diet, and exercise) on both physical and mental health. However, because Sabbath keeping was defined simply as avoidance of secular activities in the Superville study, and not in terms of the psychological experience of Sabbath keeping, the relationship between Sabbath-keeping internalization (as opposed to avoidance of behaviors) and increased well-being is as yet untested in published studies. (2)
The process of internalization is core to Self-Determination Theory (SDT; Deci & Ryan, 2008; Ryan & Deci, 2000), a theory of human motivation and behavior that construes human development and well-being as dependent on three innate psychological needs: competence (the feeling that one can do a task effectively), autonomy (the feeling of volitional control of behavior), and relatedness (feelings of interpersonal security and warmth). Importantly, SDT differentiates between motivations that are perceived as resulting from full volition and free choice (autonomous motivation) and those that are perceived as resulting from external forces (controlled motivation; Deci & Ryan, 2008; Ryan & Deci, 2000)--the same distinction that we made between those who rest on the Sabbath and those who endure it. Individuals move from controlled motivations to autonomous motivations as they first identify with the importance of an externally-controlled behavior, and then internalize and integrate it with other aspects of their selves (Deci & Ryan, 2008). This process must occur within a social context that promotes relatedness (Ryan & Deci, 2000) and provides trusted feedback about competence so that individuals know they are successfully doing those things that really matter to the community. Such autonomous behaviors are related to increased subjective well-being (Ryan & Deci, 2000).
Based on the recognition that Sabbath keeping is not always experienced as a rest, Diddams and her colleagues (2004) proposed three Sabbath-keeping models within the SDT framework that predicted corresponding differences in well-being--life segmentation, prescribed meaning, and integrated Sabbath. The life segmentation model involves attempts to place temporal boundaries between work and other important aspects of life such as family, recreation, or worship. However, Sabbath keepers following the life segmentation model face two threats to well-being: They may not be completely successful in their segmentation of time, leading to stressful rumination despite physical detachment (Cropley & Purvis, 2003), or Sabbath keeping may break down into legalistic rule-following (Diddams et al" 2004). In either case, autonomy, competence, and relatedness are threatened, and well-being is predicted to suffer. This experience is a type of partial internalization known as introjection--i.e., regulation via guilt, fear, social pressure, and shame (Neyrinck et al., 2006; Ryan et al., 1993).
The second model of Sabbath keeping--prescribed meaning--involves explicitly and intentionally associating Sabbath keeping with positive meaning (Diddams et al., 2004), but only as a means to an end. The key threat to this model of Sabbath keeping stems from perceiving oneself as hypocritical. The realization that Sabbath keeping may only occur as a means to particular ends of personal importance (e.g., time with family, future productivity, time for self-improvement), rather than as a celebration of God's covenant with humankind (Diddams et al., 2004), could lead to increased stress or to diminished perceptions of competence in religious arenas, and thus to reduced well-being.
The final model of Sabbath keeping-integrated Sabbath--involves internalizing Sabbath keeping such that the principles of Sabbath keeping are expressed throughout the entirety of the Sabbath keeper's life (Diddams et al., 2004)--not just during the Sabbath. When integrated Sabbath keepers are able to act in accordance with their selves by keeping the Sabbath and living as Sabbath keepers every day, their stress levels should be lower and their well-being higher.
Based on the integrated Sabbath model, Diddams and her colleagues (2004) suggested three ways in which Sabbath keeping could directly meet the three innate psychological needs identified by SDT (autonomy, competence, and relatedness). First, they noted that the act of resting is a statement of autonomy from external demands. Second, they suggested that completely separating from the demands of the work week is a statement of competence in the area of time management. Finally, they observed that communal Sabbath keeping builds relatedness and likely increases positive feedback, leading to further internalization of Sabbath keeping. Because of the costs and importance of Sabbath keeping in Sabbath-keeping communities, the relationship between internalization of Sabbath keeping and well-being may be stronger than the relationship previously demonstrated between low-cost practices and well-being (Neyrinck et al., 2006; Ryan et al., 1993) or between avoidance of secular activities and well-being (Superville et al., 2013). Consistent with this prediction, more orthodox Jews reported a stronger relationship between internalization and well-being than less orthodox Jews, especially in a supportive religious community setting (Zeldman, 2006).
There are three predictions that arise from the application of SDT to Sabbath keeping that will be considered in the current study. First, among Sabbath keepers, it should be possible to identify particular experiences of Sabbath keeping associated with each of the models, including the respective threats to Sabbath keeping. Second, it should be possible to identify different groups of Sabbath keepers distinguished by internalization and introjection. Finally, those Sabbath keepers who display the deepest levels of internalized Sabbath keeping should display the highest levels of subjective well-being, and the relationship should be distinct from and stronger than the relationship between widespread, low-cost Christian behaviors and well-being. The current exploratory study examines these three predictions in a large and diverse community of Sabbath keepers using the complementary techniques of factor analysis and hierarchical clustering (Gorman & Primavera, 1983).
In factor-cluster studies, factor analysis is first used to identify latent factors which can be compared to theory. The complementary hierarchical clustering analysis is meant to identify relatively homogenous groups of subjects in factor space--again for comparison to theory. We chose to use a factor-cluster approach because Diddams and her colleagues (2004) predicted both that differences in Sabbath keeping experiences would exist and that those differences would naturally sort Sabbath keepers into several groups.
We surveyed 362 subjects from a Sabbath-keeping university community about their subjective well-being and Sabbath-keeping behaviors and experiences. Of these, 10 reported that they did not keep the Sabbath and were removed from further analysis. Additionally, 27 subjects over the age of 25 were removed from the analysis, leaving 325 young adult subjects (59% female; 89% baptized church members; average age = 19.5 years; average years of Sabbath keeping = 17.3). We focused on 18- to 25-year-old subjects because young adults are in the midst of developing their own religious identities and, in this case, Sabbath-keeping practices apart from those of their parents and the religious community of their childhood (Arnett & Jensen, 2002). We therefore hoped to sample transitioning Sabbath keepers with varying degrees of internalization (homogeneity was noted among respondents in Superville et al., 2013).
All but five of the Sabbath-keeping young adults in our sample self-identified as Seventh-day Adventist Christians. The Seventh-day Adventist church traces its roots to the Second Great Awakening, a widespread Christian revival that occurred in the early 1800s in the eastern United States. As indicted by the inclusion of "Seventh-day" in the name of the denomination, Adventists identify as Christians who celebrate the Sabbath on Saturday. As a result, and of interest to this study, Sabbath keeping is normative among Adventists--between 85% and 95% of respondents in two major studies of families with at least one Adventist church member (IV = 3,069; Sahlin, 2010; Sahlin & Sahlin, 1997) reported keeping the Sabbath as a family. As the Adventist church has transitioned from sect to denomination (Cragun & Lawson, 2010), Adventist Sabbath keeping in the United States and Canada has moved from more restrictive practices meant to distinguish Adventists from other Christian churches that keep a Sunday Sabbath to greater freedom for individuals and communities to choose their own meaningful Sabbath-keeping practices while maintaining the practice of Saturday Sabbath keeping. These changes have led to an overall decrease in any form of weekly Christian Sabbath keeping in the United States and Canada (Bull & Lockhart, 2007); nevertheless, the vast majority of Adventists still punctuate their week with a 24-hour period of Sabbath keeping that is differentiated from the rest of the week. The most common Sabbath-keeping practices reported by active church members over the last two decades are attending church (> 95%), wearing "Sabbath clothes" (> 80%), preparing for Sabbath on Friday (> 70%), quiet reading (> 65%), singing or playing music (> 60%), limiting media consumption (> 50%), and walking in nature (> 50%; Sahlin, 2010; Sahlin Sc Sahlin, 1997). Because Adventists are one of the few remaining Protestant Christian groups to practice widespread, normative Sabbath keeping for a full day each week, an Adventist community was selected as the initial group with whom to test the models of Sabbath keeping. Because relatedness is a key component of one of the models proposed by Diddams and her colleagues (2004), surveying a community of Sabbath keepers was necessary.
We measured subjects' well-being using the Oxford Happiness Questionnaire (OHQ; Hills & Argyle, 2002) and the Subjective Happiness Scale (SHS; Lyubomirsky & Lepper, 1999). The OHQ is a broad but unidimensional measure of well-being--which Diddams and her colleagues (2004) used to predict a variety of well-being effects, from happiness and joy to coping and vitality. It is reliable ([alpha] = 0.91) and has good construct validity (Hills & Argyle, 2002). This scale includes a wide variety of items such as "Life is good," "I feel able to take anything on," and "I do not have fun with other people" (reverse coded). Subjects responded to the 29 items on the OHQ using a 6-point scale (strongly disagree to strongly agree)-, responses were reverse coded as appropriate (Hills & Argyle, 2002) and averaged to create a single scale score for each subject, with higher scores indicating greater well-being.
We also included the very brief four-question SHS to examine whether patterns were robust across wellbeing measures. Although it is short, the SHS has good split-half reliabilities ([alpha] = .80 to .94) and stability coefficients (r = .61 to .90) across several weeks as well as strong correlations with other measures of well-being (Lyubomirsky & Lepper, 1999). Subjects respond to the SHS on a 7-point spectrum that is unique to each question; responses to each item are then averaged to create a single scale score, with higher scores indicating greater well-being. For example, the first item on the SHS begins with the stem "In general, I consider myself..." and elicits a response on a scale anchored by "not a very happy person" and "a very happy person."
We constructed 48 initial items (3) to measure Sabbath-keeping internalization by first identifying predicted experiences (both positive and negative) related to the three models of Sabbath keeping in Diddams and colleagues' (2004) theoretical paper. We then wrote multiple items for each of the identified experiences. Sixteen of these items were taken from the life segmentation model, 19 were from prescribed meaning, and the remaining 13 were from the integrated Sabbath model. Subjects responded to the items using a 7-point Likert-type scale ranging from 1 (strongly disagree) to 7 (strongly agree). We also asked subjects 46 questions about the frequency with which they engaged in a variety of activities (both normative and non-normative among Adventists) on the Sabbath in order to characterize their compliance with Adventist Sabbath-keeping norms.
Finally, 196 of the subjects also completed the Christian Religious Internalization Scale (CRIS; Ryan et al., 1993) which has two subscales: introjection and identification. Both subscales correlate with subjective well-being (identification positively and introjection negatively) in other Judeo-Christian samples (Ryan et al., 1993; Zeldman, 2006), and have acceptable reliabilities across a variety of samples (identification: [alpha] = .69 to .82; introjection: [alpha] = .64 to .82; Ryan, et al., 1993). We included the CRIS to examine whether internalization of widespread low-cost practices strongly mediates any relationship between internalization of distinctive high-cost practices and well-being.
Subjects accessed the surveys through a research participation website as part of a course requirement (credit was assigned automatically by the system so that researchers did not know who had participated). Subjects completed an informed consent screen, followed by the surveys using a web interface on their own schedule. We felt that this was the best option as it allowed subjects to admit to non-normative Sabbath-keeping behavior (if necessary) in an anonymous context. While this means that subjects were not monitored as they completed the surveys, we have replicated the basic patterns reported in this study in two later pilot studies conducted in-person in an English-speaking church (N = 37) and a Spanish-speaking church (N = 60) as part of the process of preparing instructions for pastors on how to use the survey. Thus, we are confident that the vast majority of surveys in this study were completed honestly.
We conducted an exploratory factor analysis of the 48 Sabbath-keeping experience items using principal axis factoring with direct oblimin rotation using the fa function in the psych package version 184.108.40.206 (Revelle, 2014) in R version 3.1.1 (R Core Team, 2014). We followed best practices for factor analysis by requiring converging evidence for the number of factors to extract, including measures of internal consistency within each factor (Bandalos & Boehm-Kaufman, 2009; Costello & Osborne, 2005). We used principal axis factoring because it is robust with the highly skewed responses that were present for some of our items, and we used direct oblimin rotation because it allowed for the expected correlation between factors.
We identified an initial range of factors to retain by using converging evidence from the four heuristic tests shown to be most accurate and robust (Costello & Osborne, 2005; Velicer, Eaton, & Fava, 2000): the visual scree plot test, parallel analysis, the minimum average partial (MAP) test, and hierarchical clustering analysis (ICLUST). All of these tests were conducted using the psych package functions VSS.scree, fa.parallel, vss, and ICLUST in R 3.1.1. We examined each of the candidate factor solutions for meaningfulness in light of theory and acceptability and selected a solution for further refinement. Acceptable solutions met the criteria of each factor having at least five items loading greater than .5 and minimal cross-loading of items on multiple factors greater than .3 (Costello & Osborne, 2005).
We then iteratively refined the selected factor solution by removing redundant items with low loadings and items with minimal contributions to the factors one item at a time. The resulting factors were then named based primarily on the five highest-loading items, but in a manner consistent with all of the items loading on the factor. We sought converging evidence for the solution by again conducting the four heuristic tests for the number of factors to extract and by examining internal consistency of each factor using three measures--Cronbach's [alpha], Revehe's [beta], and McDonald's [[omega].sub.h]. While [alpha] is sensitive to the interrelatedness of items on a scale, it is insensitive to lumpiness within a scale (Schmitt, 1996), and thus overestimates--sometimes severely--the unidimensionality of a scale or subscale. Revelle's [beta] (the worst split-half reliability) is sensitive especially to one or a few problematic items, while McDonald's [[omega].sub.h] is a measure of factor saturation (Zinbarg, Revelle, Yovel, & Li, 2005).
Following the factor analysis, we clustered subjects based on their responses to the remaining items using the mclust procedure in the mclust package version 4.4 (Fraley, Raftery, & Scrucca, 2014) in R 3.1.1. Mclust allows us to compare a variety of clustering models based on the Bayesian Information Criterion (BIC) value associated with each model, and to select the best model along with an estimate of the strength of the evidence for that model (Raftery, 1995). We then examined the relationship between the resulting clusters of Sabbath keepers and their subjective well-being using a between-subjects, one-way ANOVA.
Finally, we calculated Pearson's r to quantify the magnitude and direction of the relationship between subjects' scores for each factor and their scores on the CRIS introjection and identification subscales. We then conducted a mediation analysis using the mediation function in the MBESS package version 3.3.3 (Kelley & Lai, 2012) in R 3.1.1 to examine whether any relationship between Sabbath-keeping factors and well-being was mediated by more general internalization of religious practices. We used Preacher and Kelley's (2011) [[kappa].sup.2] effect size statistic, which represents the proportion of the maximum possible indirect effect accounted for by the actual indirect effect, as a measure of mediation effect size. As a heuristic, we applied benchmarks of .01, .09, and .25 as small, medium, and large [[kappa].sup.2] effect sizes respectively for qualitative interpretation of mediation in this study (Preacher & Kelley, 2011).
The young adult subjects in this study were relatively typical Adventist Sabbath keepers. The most frequent (measured by typical number of weeks out of four) Sabbath-keeping behaviors were spending time with friends (M = 3.32), doing something that made you happy (M = 3.26), and attending a church service (M = 3.14), The least frequent behaviors were completing household chores (M = 0.65), working on a necessary school or work project (M = 0.52), and going shopping for something needed or wanted (M = 0.44).
Tests for the number of factors to retain based on the full set of items suggested four (scree plot, ICLUST) or six (MAP test, parallel analysis) factors. While the ICLUST solution suggested four clusters, three of the clusters had very few items, suggesting that only one true factor was present. We thus examined all of the factor solutions between one and seven factors.
The one- and two-factor solutions produced solid factors including most items; these solutions accounted for 35% and 41% of the variance respectively. While the five-, six-, and seven-factor solutions all accounted for over 50% of the variance, each included at least one marginal factor and several cross-loading items, an indicator of over-factoring. Likewise, the four-factor solution had three strong factors, but a marginal fourth factor and substantial cross-loading. The three-factor solution, on the other hand, explained 45% of the variance and had three strong factors--although with several cross-loadings between .3 and .4. The third factor in this solution was split in the four-factor model. While there was no theoretical support for splitting the third factor, there was theoretical support for the three-factor model over the one- and two-factor models. Thus, the three-factor solution was selected for further refinement.
A total of 15 of the 48 items were iteratively removed because they had low communalities (< 0.3; 12 items) or low factor loadings (< 0.4; 3 items). The resulting factors were unbalanced: Factors 2 and 3 had seven items each, while Factor 1 had 19 items. Thus, the 12 lowest loading redundant items were removed from Factor 1 to create three balanced factors and a shorter scale for future studies; this did not change the content of Factor 1. The resulting solution (54% of variance explained) retained three strong factors (Factor 1: 7 items, 7 loadings > 0.5, 22% of variance; Factor 2; 7 items, 7 loadings > 0.5, 16% of variance; Factor 3: 7 items, 6 loadings > 0.5, 16% of variance) with no cross-loadings.
All three factors had high internal reliability and acceptable measures of unidimensionality (see Table 1). The four heuristics for number of factors centered on the three-factor solution: the scree plot suggested two factors, MAP and ICLUST suggested three, and parallel analysis suggested four. The ICLUST solution, however, suggested different groupings of items than the factor analysis. While this could be evidence of an over-factored solution (an interpretation belied by the measures of subscale unidimensionality), we believe that this is evidence that there is a hierarchical relationship between the subscales and the full scale. Further evidence for a hierarchy comes from measures of full scale reliability and unidimensionality: Cronbach's [alpha] for the full scale was .93; Revelle's [beta] was .74; and MacDonald's [[omega].sub.h] was .76, suggesting that the subscales could also be interpreted as a single scale. Of course, a hierarchical relationship is not surprising, as theory (Diddams et al., 2004) posits just such a relationship. There was no evidence of underfactoring as four- and five-factor solutions yielded weak factors that split one of the strong factors from the three-factor solution.
The three-factor solution (see Table 1 for factor structure and the Appendix for the final set of items) supported the proposed models underlying internalization of Sabbath keeping. We therefore named the factors in accordance with the theoretical proposal (Diddams et al., 2004). Factor 1 was named prescribed meaning and included items related to perceived personal benefits of Sabbath keeping; higher scores indicated higher perceived personal benefits. Factor 2 was named incomplete segmentation and included items related to the stress of keeping the Sabbath or the imposition of Sabbath keeping on personal goals and time; higher scores indicated greater perception of imposition. The related model was named life segmentation in the theoretical proposal, but it was the threats to or negative aspects of Sabbath keeping which loaded strongly on this factor, and thus motivated the name change. Factor 3 was named integrated Sabbath and was related to the experience of Sabbath keeping as a core part of a person's life; higher scores on this factor indicated higher levels of integration. As expected, the factors were highly related: prescribed meaning and integrated Sabbath were positively correlated (r = .69), and incomplete segmentation was negatively correlated with both prescribed meaning (r = -.54) and integrated Sabbath (r = -.55).
The best clustering solution using the 21 items had three clusters and a BIC of -20671.41 and was highly interpretable. The next best fitting model had two clusters and a BIC of -20803.21. The difference in BIC between the two best fitting models was 131.80, which is considered to be "very strong" evidence for the three-cluster solution according to Raftery's (1995) criteria.
The three clusters were named using constructs from SDT based on the pattern of average scores on the three factors identified in the factor analysis (see Table 2). Subjects in Cluster 1, introjected Sabbath keepers, showed moderate scores on the integrated Sabbath and prescribed meaning subscales and higher scores on the incomplete segmentation subscale relative to other clusters. The introjected Sabbath keepers contrasted with Cluster 2, identified Sabbath keepers, and Cluster 3, integrated Sabbath keepers, in which subjects showed significantly lower scores on the incomplete segmentation subscale, F(2, 322) = 256.8, p < .001, and significantly higher scores on both the prescribed meaning, F(2, 322) = 203.4, p < .001, and the integrated Sabbath subscales, F(2, 322) = 232.5, p < .001. Similar patterns were seen between clusters on the CRIS. Integrated Sabbath keepers scored highest on the identified regulation subscale, followed by identified and then introjected Sabbath keepers, F(2, 193) = 11.9, p < .001; responses to the introjected regulation subscale showed the inverse pattern, F(2,193) = 30.8, p < .001.
Finally, subjects in the three clusters reported levels of subjective well-being that paralleled the differences in internalization of Sabbath keeping. Based on the responses to both the OHQ, F(2, 322) = 10.3, p < .001, and the SHS, F(2, 322) = 34.5, p < .001, integrated Sabbath keepers had the highest well-being scores, followed by identified Sabbath keepers, and then introjected Sabbath keepers. This pattern was not simply an artifact of the clustering procedure. Incomplete segmentation was negatively correlated with well-being (OHQ: r = -.37, p < .001; SHS: r = -.16, p < .001), while prescribed meaning (OHQ: r = .41, p < .001; SHS: r = .21, p < .001) and integrated Sabbath (OHQ: r = .41, p < .001; SHS: r = .28, p < .001) were positively correlated.
The mediation analysis used the OHQ measure of well-being because of its increased length and breadth of conceptual coverage and the total Sabbath keeping scale scores for the same reason; however, the patterns described for the total scale (including all 21 items) are representative of the patterns for the three subscales.
Christian religious identification (r = .38, p < .001) and introjection (r = -.24, p < .01) both correlated with OHQ well-being but only weakly mediated the relationship between Sabbath keeping internalization and well-being. The introjection model accounted for about one fifth of the variance in well-being, [R.sup.2] = .212, 95% CI [.112, .313]; however, the mediation effect size was trivial to small, [[kappa].sup.2] = .034; bootstrapped 95% CI [.002, .084]. The identification model accounted for a similar proportion of variance, [R.sup.2] = .231, 95% CI [.128, .333]; the mediation effect size was small to medium, [[kappa].sup.2] = 0.099; bootstrapped 95% CI [.031, .173].
Is internalization of a meaningful and costly practice--Sabbath keeping--within a religious community related to increased well-being? We found that self-reported Sabbath-keeping experiences did indeed fit the three proposed Sabbath-keeping models (Diddams et al., 2004). Consistent with SDT, items representing the incomplete segmentation model were related to controlled regulation of Sabbath keeping while items representing the prescribed meaning and integrated Sabbath models were related to autonomous regulation. We also identified three Clusters of Sabbath keepers who differed in internalization and autonomous regulation--again consistent with the predictions of Diddams and her colleagues (2004). In both correlational and cluster-comparison analyses, we found that deeper Sabbath-keeping internalization was related to higher levels of subjective well-being but was only weakly mediated by internalization of low-cost Christian behaviors, even though the latter also correlated significantly with subjective well-being. Thus, internalization of meaningful high-cost practices is related to increases in well-being independent of internalization of low-cost practices.
Within the SDT framework, this makes sense, as the signals of autonomy, competence, and relatedness that matter for any given believer will depend on their understanding of what religious practices matter most within their religious community (Diener et al., 2011; Neyrinck et al., 2006, Xygalatas et al., 2013). Moreover, contextualizing religious internalization opens the door for a richer account of why the same religious practices may lead to increased well-being for some believers and decreased well-being for others (Mochon et al., 2011). Indeed, the curvilinear relationship between well-being and commitment (Mochon et al., 2011) and the relationship between decreasing commonality with other believers and disaffiliation (Vargas, 2012) are both predicted by an internalization of distinctive costly practices account. Likewise, these results suggest a psychological component to the sociological observation that conservative churches that place substantial demands on members (including the Seventh-day Adventist church; Iannaconne, 1994; Kelley, 1972) continue to outpace liberal and mainline churches in growth. However, the current study is directly applicable only to the Seventh-day Adventist church--and only the church in the United States at that; this suggests the need for further study of distinctive high-cost practices among other strict religious groups (Xygalatas et al., 2013). Indeed, the unique contribution of distinctive religious practices to well-being may indicate that internalization of religion is best understood within particular religious contexts that share features (such as high-cost demands) but express those shared features in denominationally-specific ways.
What is absent from the current study, of course, is evidence that the internalization of distinct practices relates to well-being through the mechanism of need support as proposed in SDT. Future research on the internalization of distinctive religious practices should identify the degree to which the basic psychological needs of autonomy, competence, and relatedness are related to internalization of those practices and increased well-being. This study is also correlational--the possibility still remains that well-being is driving internalization of Sabbath-keeping, or (most likely) that the two amplify each other. Our focus on young adults also limits generalizability; however, we have found the same relationships (although slightly attenuated) between the three subscales and well-being in pilot studies with older adult samples. Finally, this study, like any factor-analytic study, is limited by the initial pool of items. There may be aspects of Sabbath-keeping internalization that we cannot see because we had no items to query those aspects. This calls for qualitative, anthropological, and mixed-methods approaches that can explore the phenomenology of Sabbath-keeping internalization.
In his 2014 book, Sabbath as Resistance, Walter Brueggeman develops a theology of the Sabbath as an intentional antidote to the many aspects of life that tear at our relationships with each other and our God. What our study suggests is that implementing Sabbath keeping as a protection from the "cares of this world" (Mk. 4:1-20) will demand not only personal action, but the support of a committed community that places value on the cost of Sabbath keeping and buffers that cost through social support. Moreover, this study predicts that other costly and meaningful practices (religious or secular) that, like Sabbath keeping, build relatedness, competence, and autonomy might also have similar effects, but only when followed strictly. Identifying such practices can focus churches and believers on those actions that, if internalized, can allow life to be lived "more abundantly" (John 10:10).
Karl G. D. Bailey and Arian C. B. Timoti
Author Note: The authors would like to thank Chinyere Sampson and Cheryl B. Simpson for helpful discussions during survey development and for assistance with data collection. We would also like to thank L. Monique Pittman, Ante Jeroncic, Vanessa Corredera, and Charles Abreu for their many helpful comments on drafts of this manuscript.
Correspondence concerning this article should be addressed to Karl Bailey, Behavioral Sciences Department, 8488 E. Campus Circle Drive, Andrews University, Berrien Springs, MI 49104-0030.
Arnett, J.J., & Jensen, L. A. (2002). A congregation of one: Individualized religious beliefs among emerging adults. Journal of Adolescent Research,, 17,451-467. doi:10.1177/0743558402175002
Bandalos, D. L., & Boehm-Kaufman, M. R. (2009). Common misconceptions in exploratory factor analysis. In C. E. Lance and R. J. Vandenberg (Eds.), Statistical and methodological myths and urban legends: Where pray tell did they get this idea? (pp. 63-88). New York, NY: Routledge.
Bass, D. C. (1997). Keeping Sabbath: Reviving a Christian practice. Christian Century, 114(1), 12-16.
Brueggemann, W. (2014). Sabbath as resistance: Saying no to the culture of now. Louisville, KY: Westminster John Knox Press.
Bull, M. & Lockhart, K. (2007). Seeking a sanctuary: Seventh-day Adventism and the American dream. Bloomington, IN: Indiana University Press.
Costello, A. B., & Osborne, J. W. (2005). Best practices in exploratory factor analysis: Four recommendations for getting the most from your analysis. Practical Assessment, Research & Evaluation, 10(7). Retrieved from http://pareonline.net/pdf/v10n7.pdf
Cragun, R. T., & Lawson, R. (2010). The secular transition: The worldwide growth of Mormons, Jehovah's Witnesses, and Seventh-day Adventists. Sociology of Religion, 71, 349-373. doi:10.1093/ socrel/srq022
Cropley, M., & Purvis, L. J. M. (2003). Job strain and rumination about work issues during leisure time: A diary study. European Journal of Work and Organizational Psychology, 12, 195-207. doi:10.1080/13594320344000093
Deci, E. L., & Ryan, R. M. (2008). Facilitating optimal motivation and psychological well-being across life's domains. Canadian Psychology, 49, 14-23. doi:10.1037/0708-55220.127.116.11
Diddams, M., Surdyk, L. K., & Daniels, D. (2004). Rediscovering models of Sabbath keeping: Implications for psychological wellbeing .Journal of Psychology and Theology, 32, 3-11.
Diener, E., Tay, L., & Myers, D. G. (2011). The religion paradox: If religion makes people happy, why are so many dropping out? Journal of Personality and Social Psychology, 101, 1278-1290. doi:10.1037/ a0024402
Dudley, R. L. (1978). Alienation from religion in adolescents from fundamentalist religious homes. Journal for Scientific Study of Religion, 17, 389-398.
Fraley, C., Raftery, A., & Scrucca, L. (2014). mclust: Normal mixture modeling for model-based clustering classification, and density estimation. (R package Version 4.4) [Software]. Seattle: University of Washington. Available at http://cran.r-project.org/web/packages/ mclust/
Gorman, B. S., & Primavera, L. H. (1983). The complementary use of cluster and factor analysis methods .Journal of Experimental Education, SI, 165-168.
Hackney, C. H., & Sanders, G. S. (2003). Religiosity and mental health: A meta- analysis of recent studies. Journal for the Scientific Study of Religion, 42, 43-55. doi:10.1111/1468-5906.t01-1-00160
Hills, P., & Argyle, M (2002). The Oxford Happiness Questionnaire: A compact scale for the measurement of psychological wellbeing. Personality and Individual Differences, 33, 1073-1082. doi:10.1016/S0191-8869(01)00213-6
Hornsey, M. J., & Jetten, J. (2004). The individual within the group: Balancing the need to belong with the need to be different. Personality and Social Psychology Review, 8, 248-264. doi:0.1207/ s15327957pspr0803_2
Iannaconne, L. R. (1994). Why strict churches are strong. American Journal of Sociology, 99, 1180-1211. doi:10.1086/230409
Kasser, T., Rosenblum, K. L., Sameroff, A. J., Deci, E. L., Niemiec, C. P., Ryan, R. M., ... Hawks, S. (2014). Changes in materialism, changes in psychological well-being: Evidence from three longitudinal studies and an intervention experiment. Motivation and Emotion, 38, 1-22. doi:10.1007/s11031-013-9371-4
Kelley, D. M. (1972). Why conservative churches are growing: A study in the sociology of religion. New York, NY: Harper & Row.
Kelley, K., & Lai, K. (2012). MBESS. (R package Version 3.3.3) [Software]. South Bend: Notre Dame University. Available at http://cran.r-project.org/web/packages/MBESS/
Lyubomirsky, S., & Lepper, H. S. (1999). A measure of subjective happiness: Preliminary reliability and construct validation. Social Indicators Research, 46, 137-155.
Maltby, J., & Day, L. (2003). Religious orientation, religious coping and appraisals of stress: Assessing primary appraisal factors in the relationship between religiosity and psychological well-being. Personality and Individual Differences, 34, 1209-1224. doi:10.1016/ S0191-8869(02)00110-1
Mochon, D., Norton, M. I., & Ariely, D. (2011). Who benefits from religion? Social Indicators Research, 101, 1-15. doi:10.1007/s11205010-9637-0
Myers, D. G. (2000). The funds, friends, and faith of happy people. American Psychologist, 55, 56-67. doi:10.1037//0003-066X.55.1.56
Neyrinck, B., Vansteenkiste, M., Lens, W., Duriez, B., & Hutsebaut, D. (2006). Cognitive, affective and behavioral correlates of internalization of regulations for religious activities. Motivation and Emotion, 30, 323-334. doi:10.1007/sl 1031-006-9048-3
Pichon, I., Boccato, G., & Saroglou, V. (2007). Nonconscious influences of religion on prosociality: A priming study. European Journal of Social Psychology, 37, 1032-1045. doi:10.1002/ejsp.416
Preacher, K. J., & Kelley, K. (2011). Effect size measures for mediation models: Quantitative strategies for communicating indirect effects. Psychological Methods, 16, 93-115. doi:10.1037/a0022658
R Core Team (2014). R: A language and environment for statistical computing. (Version 3.1.1) [Software] Vienna: R Foundation for Statistical Computing. Available at http://www.r-project.org/
Raftery, A. E. (1995). Bayesian model selection in social research. Sociological Methodology, 25, 111-163. doi:10.2307/271063
Revelle, W. (2014). psych: Procedures for psychological, psychometric, and personality research. (R package Version 18.104.22.168) [Software]. Evanston: Northwestern University. Available at http://cran.rproject.org/web/packages/psych/
Ryan, R M., & Deci, E. L. (2000). Self-determination theory and the facilitation of intrinsic motivation, social development, and well-being. American Psychologist, 55, 68-78. doi:10.1037/0003066X.55.1.68
Ryan, R. M., Rigby, S., & King, K. (1993). Two types of religious internalization and their relations to religious orientations and mental health. Journal of Personality and Social Psychology, 65, 586-596. doi:10.1037/0022-3522.214.171.1246
Sahlin, M. (2010). Adventist families in North America. Milton-Freewater, OR: Center for Creative Ministry.
Sahlin, M., & Sahlin, N. (1997). A new generation of Adventist-families: What it means to you and your church. Portland, OR: Center for Creative Ministry.
Scheitle, C. P., & Adamczyk, A. (2010). High-cost religion, religious switching, and health. Journal of Health and Social Behavior, 51, 325-342. doi:10.1177/0022146510378236
Schmitt, N. (1996). Uses and abuses of coefficient alpha. Psychological Assessment, 8, 350-353. doi:10.1037/1040-35126.96.36.1990
Sosis, R, & Bressler, E. R (2003). Cooperation and commune longevity: A test of the costly signaling theory of religion. Cross-Cultural Research, 37, 211-239. doi:10.1177/1069397103037002003
Superville, D. J., Pargament, K. I., & Lee, J. W. (2013). Sabbath keeping and its relationships to health and well-being: A mediational analysis. InternationalJournal for the Psychology of Religion, 24, 241-256. doi:10.1080/10508619.2013.837655
Thomas, J. N., & Olson, D. V. A. (2010). Testing the strictness thesis and competing theories of congregational growth. Journal for the Scientific Study of Religion, 49, 619-639. doi:10.1111/j.14685906.2010.01534.x
Vargas, N. (2012). Retrospective accounts of religious disaffiliation in the United States: Stressors, skepticism, and political factors. Sociology of Religion, 73, 200-223. doi:10.1093/socrel/srr044
Velicer, W. F., Eaton, C. A., & Fava, J. L. (2000). Construct explication through factor or component analysis: A review and evaluation of alternative procedures for determining the number of factors or components. In R D. Goffin & E. Helmes (Eds.), Problems and solutions in human assessment: Honoring Douglas N. Jackson at seventy (pp. 41-71). Boston, MA: Kluwer.
Waller, S. N. (2009). Favourite pew or box seat? Sabbath beliefs as a barrier to sporting event attendance on Sunday: A congregational study .Journal of Religion and Popular Culture, 21, 1-15. doi:10.3138/jrpc.21.2.005
Weinstein, N., & Ryan, R. M. (2011). A self-determination theory approach to understanding stress incursion and responses. Stress and Health, 27, 4-17. doi:10.1002/smi. 1368
Xygalatas, D., Mitkidis, P., Fischer, R., Reddish, P., Skewes, J., Geertz, A. W., ... Bulbulia, J. (2013). Extreme rituals promote prosociality. Psychological Science, 24, 602-1605. doi:10.1177/0956797612472910
Zeldman, A. (2006). Religion and mental health: Correlates in internalization, centrality, and God representations. (Unpublished doctoral dissertation). University of Rochester, Rochester, NY.
Zinbarg, R E., Revelle, W., Yovel, I., & Li, W. (2005). Cronbach's a, Revelle's [beta], and McDonald's [[omega].sub.h]: Their relations with each other and two alternative conceptualizations of reliability. Psychometrika, 70, 123-133. doi:10.1007/s11336-003-0974-7
(1) The authors would like to thank an anonymous reviewer for noting this relationship.
(2) Superville and his colleagues do allude to preliminary findings consistent with internalization.
(3) The full set of initial items is available from the first author.References
BAILEY, KARL G. D. PhD. Address: Behavioral Sciences Department, Andrews University, 8488 E Campus Circle Drive, Berrien Springs, MI 49104. Title: Professor of Psychology. Degrees: BS (Biology and Psychology) Andrews University; MA (Psychology) Michigan State University; PhD (Psychology) Michigan State University. Specializations: Psychology of motivation, visual cognition, and psycholinguistics.
TIMOTI, ARLAN C. B. MS. Address: Behavioral Sciences Department, Andrews University, 8488 E Campus Circle Drive, Berrien Springs, MI 49104. Title: Instructor. Degrees: BS (Biology) Oakwood University; MS (Community & International Development) Andrews University. Specializations: Non-profit management, development education.
Appendix Final set of items included in the Sabbath Keeping Inventory Item # Subscale Item Text 1 INT Nobody makes me keep the Sabbath--it is just a part of how I live. 2 PRE I am healthier because I keep the Sabbath. 3 PRE I use the Sabbath to spend time with people who are important to me. 4 PRE Keeping the Sabbath reduces my stress. 5 INC I keep the Sabbath because other people are watching what I do. 6 INC If I didn't keep the Sabbath, I would get a lot more done. 7 PRE Keeping the Sabbath helps me understand what is important to me. 8 PRE When I have real Sabbath rest, I cope better with the stresses of the week. 9 INT Keeping the Sabbath is a way of life, not just something that happens once a week. 10 INC Following Sabbath rules brings more stress to me. 11 INT Keeping the Sabbath is an intentional act on my part. 12 INT Keeping the Sabbath helps me to discover more about who I am. 13 INC Keeping the Sabbath prevents me from doing what needs to be done. 14 INT I grow the most as a person because of how I spend my Sabbaths. 15 INC Sabbath is very stressful. 16 INC Even when I keep the Sabbath, I keep thinking about all of the other things that I need to get done. 17 INT Keeping the Sabbath is part of who I am, not what I do. 18 PRE The Sabbath is an opportunity to reconnect with friends that I've missed throughout the week. 19 INT I apply what I learn on Sabbath to how I live during the rest of the week. 20 INC I find it stressful to be forced to take a break on Sabbath from what I need to get done. 21 PRE Keeping the Sabbath helps me redefine what is important to me. Note. Average the scores within each subscale to calculate the three subscale scores. Reverse code the incomplete segmentation (INC) items when calculating a full scale score. INC = incomplete segmentation; PRE--prescribed meaning; INT = integrated Sabbath. TABLE 1 Items, loadings, communalities, and measures of internal consistency for the three-factor solution Incomplete Prescribed Item Segmentation Meaning I would get a lot more done .70 -.02 forced to take a break .69 -.11 prevents doing what needs to be done .67 .00 keep thinking about other things .65 -.02 following Sabbath rules brings stress .57 -.12 Sabbath is very stressful .57 -.22 other people are watching .56 .08 cope better with weekly stresses -.04 .88 helps understand what is important .03 .80 helps redefine what is important .02 .77 reduces my stress -.18 .72 healthier because I keep the Sabbath -.13 .72 reconnect with friends .09 .70 spend time with people .01 .56 part of who I am, not what I do -.11 -.05 Sabbath is part of how I live -.19 -.02 I apply to how I live .17 .11 Sabbath is a way of life -.11 .10 helps me discover who I am .07 .20 an intentional act on my part -.22 .06 grow as a person .21 .27 Internal Consistency Cronbach's [alpha] .87 .91 Revelle's [beta] .74 .72 McDonald's [[omega].sub.h] .75 .83 Integrated Item Sabbath [h.sup.2] I would get a lot more done .00 .51 forced to take a break -.05 .61 prevents doing what needs to be done -.04 .48 keep thinking about other things .05 .41 following Sabbath rules brings stress -.17 .56 Sabbath is very stressful .03 .46 other people are watching -.27 .47 cope better with weekly stresses -.05 .75 helps understand what is important .15 .79 helps redefine what is important .15 .75 reduces my stress -.03 .63 healthier because I keep the Sabbath -.03 .59 reconnect with friends -.07 .39 spend time with people .05 .35 part of who I am, not what I do .71 .55 Sabbath is part of how I live .69 .62 I apply to how I live .63 .41 Sabbath is a way of life .60 .53 helps me discover who I am .60 .52 an intentional act on my part .54 .53 grow as a person .45 .34 Internal Consistency Cronbach's [alpha] .86 Revelle's [beta] .77 McDonald's [[omega].sub.h] .78 Note. Results are from an exploratory principal axis factor analysis with direct oblimin rotation. Only the 21 retained items are shown in shortened form. TABLE 2 Average responses by cluster to Sabbath--keeping subscales, measures of well-being, internalization of Christian behaviors, years of Sabbath keeping, and age Introjected Identified Sabbath Sabbath Keepers Keepers N 89 178 Age 19.37 (1.43) 19.49 (1.53) Years Keeping the Sabbath 16.26 (5.26) 17.65 (4.87) Incomplete Segmentation 3.70 (0.95) 1.94 (0.67) Prescribed Meaning 4.67 (0.96) 6.09 (0.59) Integrated Sabbath 4.37 (0.76) 5.64 (0.61) CRIS: Introjected Regulation 3.49 (1.07) 3.07 (0.99) CRIS: Identified Regulation 4.73 (0.86) 5.55 (0.84) Oxford Happiness 3.77 (0.52) 4.22 (0.57) Subjective Happiness Scale 4.82 (1.11) 5.11 (1.07) Integrated Sabbath Keepers N 58 Age 19.55 (1.53) Years Keeping the Sabbath 17.84 (4.15) Incomplete Segmentation 1.25 (0.29) Prescribed Meaning 6.80 (0.27) Integrated Sabbath 6.55 (0.36) CRIS: Introjected Regulation 2.46 (0.86) CRIS: Identified Regulation 6.01 (0.58) Oxford Happiness 4.47 (0.45) Subjective Happiness Scale 5.63 (0.90) Note. Standard deviations are in parentheses.
|Printer friendly Cite/link Email Feedback|
|Author:||Bailey, Karl G.D.; Timoti, Arian C.B.|
|Publication:||Journal of Psychology and Theology|
|Date:||Sep 22, 2015|
|Previous Article:||Relational Integration, Part II: relational integration as developmental and intercultural.|
|Next Article:||The influence of affective priming on one's perceived relationship to God.|