1. Background
It is not surprising that the concept God is pre-eminent in National
Interfaith Coalition on Aging’s [italics for emphasis] description of spiritual well-being as ‘the affirmation of life in relationship with God, self, community and environment that nurtures and celebrates wholeness’ [
1]. By 1983, a broad understanding of spirituality had been unleashed from its roots in Roman Catholic monastic moorings [
2], which provided a traditional bi-polar view of spirituality as: (i) the individual on a quest; (ii) to find her/his truest self in the context of reality [
3]. This reality, or cosmic totality, may be God or some other Transcendent, a generic, secular term that had replaced God for some people by the end of the twentieth century [
4]. A more recent shift in the understanding of spirituality was not instigated by the further inclusion of secular ideas into definitions of spirituality, but by a move from the dual structure (of personal quest, for an external reality) to an inward focus. The notion of the Self became paramount, with individual freedom and search for meaning becoming the whole focus of spirituality for many [
4]. This focus is evident in the structure of numerous instruments designed to measure spirituality and spiritual health/well-being investigated by Fisher [
5]. The most significant focus of these measures was on relationship with self, exceeding that of relationship with God, followed by relationship with others, with least emphasis on the environment.
Very few instruments contain equal numbers of items by which to assess these four domains of spirituality or spiritual health/well-being, namely relationships with self, others, environment and Transcendent Other (commonly called God). The Spiritual Health And Life-Orientation Measure (SHALOM), which was first reported in 1999 [
6], is one such instrument. It has been sought for use in over 200 studies in 20 languages [
7]. SHALOM has been thoroughly tested for its statistical validity [
8,
9,
10]. Its 20 items are distributed evenly over four domains of spiritual well-being, and measures quality of relationships with oneself, with others, with the environment and/or with a Transcendent Other [
7]. Simple language was selected to enable respondents to understand and probe complex constructs, as SHALOM was developed for use with secondary school students, in the belief that an instrument suitable for them would also be useful for adults. Consequently, the words ‘God’, ‘Divine’ and ‘Creator’ appeared in four of the five items used to assess the Transcendental domain of spiritual well-being, because notions such as ‘higher power’ and ‘godlike force’ were less comprehensible to the students [
11].
Over the last nineteen years, participants in studies using SHALOM have made negligible comment on the instrument. However, a few challenges have been presented from a small minority, during the author’s 46 conference presentations and workshops on spiritual well-being, in Australia and overseas, almost exclusively with westerners, except for work in Hong Kong. It is mainly ‘academics’ who have voiced concern that the current form of SHALOM favors people with theistic beliefs, because it includes the word ‘God.’ This criticism is not valid, because SHALOM employs a double-response method of comparing each person’s ideals with their lived experience. So, if a person does not embrace a belief in any god, then there will be no dissonance when they score ‘very low’ (equivalent to negligible or none) on lived experience in this domain, as it will match their presumably ‘very low’ scores for their ideals relating to spiritual well-being. This function separates SHALOM from other instruments which can underrate people’s spiritual well-being if they score low on lived experience in this area, in comparison with other people’s scores.
Three Australian converts to Buddhism and one American philosophy lecturer in Hong Kong questioned the use of the word ‘God.’ One American post-graduate student refused to look closely at SHALOM, and two more were refused permission to use it in their research, because of the God-word. Opposition by atheists on research and ethics committees had to be overcome on four occasions at the University of Ballarat, and Ballarat Health Services, in their attempts to eliminate the word ‘God.’ Ballarat is a regional city of just under 100,000 people. Universities and Health Services in nine larger Australian cities, as well as other American studies, have had no problems permitting the use of SHALOM for research. The word ‘God’ evokes emotive responses in many people, to the extent that one nurse educator in Ballarat tried to insist that it be written with a small ‘g’.
Following a public lecture in Hong Kong, discussion with the Director of the Centre for Religious and Spirituality Education led to the idea of using the word ‘Transcendent,’ with a variety of meanings for people to choose from, as an alternative to ‘God,’ to see if it would better accommodate views held by adherents of religious traditions as well as non-religious. The word ‘deity’ had been used in a Chinese translation of SHALOM [
12,
13], although previous research with an English version of SHALOM among predominantly Buddhist Master of Nursing students in Hong Kong had not created any comment [
14]. Subsequent email communication with the philosophy lecturer in Hong Kong placated his concerns when he was advised of the use of the word ‘Transcendent.’ A research student in Puerto Rico was also advised to change the word God to Transcendent in SHALOM, and provide six alternatives. This led to acceptance by her supervisors of her project, which they had previously opposed because of the God-word, even though Puerto Rico reports 95% with theistic beliefs and only 2.3% non-religious in its population [
15].
The most contentious issue in discussions of spiritual well-being relates to the question of a Transcendent Other (that is, some-One or some-Thing that exists beyond the human and natural worlds) having influence over humans. Worldviews and opinions are many, but empirical studies of them are few. Research needs to focus clearly on discussions of the Transcendent. To this end, the project being reported in this paper investigated whether replacing the words ‘God,’ ‘Divine’ and ‘Creator’ with the word ‘Transcendent’ would lead to as valid a spiritual well-being questionnaire as the original SHALOM has proven to be. If so, the modified SHALOM could be a more generic instrument, suitable for use with people embracing worldviews ranging from belief in God to disbelief in any Transcendent. Such a generic instrument would help overcome the difficulty many researchers have of ensuring they canvas all world-views impartially, without appearing to favor, or in fact favoring, one over others.
3. Results and Discussion
3.1. Participants
A total of 409 completed web-surveys were received from people in 14 geographic regions (Australia/New Zealand 62, Pacific OECD 3, Central Asia/China 24, Southern Asia 37, Middle East/North Africa 9, Saharan Africa 8, Southern Europe 5, Central Europe 31, Western Europe 46, former Soviet Union 51, UK 56, North America 69, Central America 1, South America 7). Fifty six percent of these were in direct response to a written invitation from the author, 40% by referred invitation, and 4% by direct access to the survey on the web. Females comprised 59% of the cohort (mean age 38.8 years, SD = 15.1), with 41% males (mean age 45.7 years, SD = 17.1). English was not the primary language spoken for 52.3% of the respondents. Sixty-five percent were either married (58%) or had been married (6% divorced/separated, 1% widowed).
The education level of the respondents exceeded that of the general population with 64% claiming a post-graduate degree, 8% a post-graduate diploma and 16% an undergraduate degree. The majority were professionals working with people (78%) and technology (9%) or were managers (8%). A wide range of worldviews was held by participants (agnostic 21, atheist 26, Baha’i 1, Buddhism 28, Catholic/Orthodox Christianity 110, Protestant Christianity 137, Christian Science 10, environmentalism 6, existentialism 5, Hari Krishna 1, Hindu 5, humanism 9, Islam 24, Judaism 10, Jehovah’s Witness 1, native spirituality 4, Shinto 2, Tao 5, Wicca 2, Zoroastrianism 1). In short 18% claimed no religious affiliation, 27% Catholic, 33% Protestant, 22% other religions. A broad range of views was also shown on the description of a Transcendent influencing the participants (Allah 23, Angel/s 11, Buddha 6, Deceased person 2, Deity/deities 5, the Divine 11, Fate 21, Father God 58, Gaia 1, God 150, Higher power 23, Higher Self 9, Mystery 10, Otherness 5, Presence 13, Something there 6, Universe/universal spirit 16; Not an area believed in 39).
3.2. Comparison Groups
As the respondents to the web-survey were more educated than the general population, a similarly-sized sample of university students and staff was selected from respondents in previous studies, which used the original SHALOM [
22,
23]. This sample was made up of 378 older students and staff in an Australian public university (see
Table 1 for details). Another sample, comprised of 460 secondary school students, aged 12 to 18 years (mean = 15.3, SD = 1.7), among whom SHALOM was originally developed, in Years 7 to 12 in State, Catholic, Christian Community and other independent schools in Victoria, Australia [
11].
Table 1.
Distribution of samples in three studies with the SHALOM.
Table 1.
Distribution of samples in three studies with the SHALOM.
| | F | | Age | % | | | Religious affiliation % |
Group | n | % | <20 | 20s | 30s | 40s | 50+ | none | Cath. | Prot. | other |
Sec school | 460 | 49 | 100 | | | | | 22 | 26 | 50 | 2 |
Uni stu staff | 378 | 72 | | 53 | 15 | 18 | 13 | 40 | 27 | 30 | 3 |
Web survey | 409 | 59 | 3 | 27 | 18 | 16 | 36 | 18 | 27 | 33 | 12 |
3.3. Psychometric Properties of the Revised, Generic SHALOM, and Comparison with the Original
Principal component analysis (PCA) of responses from the web survey of the lived experience component of the revised SHALOM revealed the presence of four components, which explained 72.5% of total variance. All the variables loaded onto the appropriate scales revealing sound measures for Personal, Communal, Environmental and (modified) Transcendental domains of spiritual well-being. Reliability analyses of the scales yielded good Cronbach’s á-coefficients, with the Transcendental domain yielding a very good value of .95 (for details see
Table 2).
When subjected to principal components analysis, these four domains cohered into a single higher-order factor, called spiritual well-being (SWB) (Kaiser-Meyer-Olkin value = .74, 63.8% total variance explained). As spiritual well-being is made up of these four domains, linear regression analysis was performed to determine the relative contribution made by each of them to the whole. At .42, the â-value for Transcendental SWB far exceeded the others (Personal SWB .27, Communal SWB .25, Environmental SWB .33), indicating that Transcendental spiritual well-being explained greatest variance in spiritual well-being overall. This finding parallels those of the original SHALOM, which concluded that relating with God is shown to be most important for spiritual well-being [
11,
24].
Confirmatory factor analyses (CFA) were performed using AMOS to investigate the relative construct validity of the original and modified versions of SHALOM, with their four component factors. As there is not a sole appropriate test, a range of fit indices was selected to test the CFA models to see which best represents the data sets. Absolute fit indices were the relative Chi-square (×
2/df), Akaike’s Information Criterion (AIC), and the Expected Cross-Validation Index (ECVI). Relative Fit Indices were the Incremental Fit Index (IFI), Normed Fit Index (NFI) and Tucker-Lewis Index (TLI). The Parsimonious Normed Fit Index (PNFI) was used. Noncentrality-based Indices were the Root Mean Square Error Approximation (RMSEA) and the Comparative Fit Index (CFI). The target values to indicate good fit indices are listed in the bottom row of
Table 3 [
25,
26].
Table 2.
Pattern matrix for PCA with Oblimin Rotation of 4-factor solution of generic SHALOM items, together with Cronbach’s alpha coefficient for reliability analyses of scales.
Table 2.
Pattern matrix for PCA with Oblimin Rotation of 4-factor solution of generic SHALOM items, together with Cronbach’s alpha coefficient for reliability analyses of scales.
Item | Com SWB | Tra SWB | Env SWB | Per SWB |
Identity | | | | .88 |
Self-awareness | | | | .82 |
Joy | | | | .40 |
Inner peace | | | | .41 |
Meaning | | | | .40 |
Love others | .77 | | | |
Forgive others | .79 | | | |
Trust | .68 | | | |
Respect others | .79 | | | |
Kindness | .83 | | | |
Connect with nature | | | .79 | |
Awe at view | | | .62 | |
One with nature | | | .92 | |
Harmony with environment | | | .79 | |
‘magic’ in environment | | | .80 | |
Relation with Transcendent | | .92 | | |
Worship Transcendent | | .90 | | |
One with Transcendent | | .86 | | |
Peace with Transcendent | | .89 | | |
Prayer life | | .88 | | |
Eigenvalue (% variance) | 8.93 (44.7%) | 2.78 (13.9%) | 1.87 (9.3%) | 1.00 (4.6%) |
alpha value | .88 | .95 | .87 | .86 |
mean(SD) | 4.02(.74) | 3.33(1.25) | 3.32(.98) | 3.98(.78) |
Values of fit indices for the lived experience component of the modified version of SHALOM displayed borderline acceptability, for relative Chi-square (with ×
2/df > 3.5, where values below 2 are good, those between 2 and 3 are generally considered acceptable, with some people accepting up to 5) and RMSEA (value of .078, which is just below the normally accepted upper limit of .08). The two cohorts that were tested using the original version of SHALOM, employing the term ‘God’, displayed acceptable to good values on the fit indices investigated. Despite considerable differences in age and world-views, especially religious aspects, between the secondary school and university cohorts studied, their fit indices for the original SHALOM were markedly similar, indicating the consistency of this instrument as a measure of spiritual well-being. Although the modified, more generic form of SHALOM, using the term Transcendent, is an acceptable measure of spiritual well-being, it is not as good as the original, which employs the term God in the measure of Transcendental SWB (see
Table 3).
Table 3.
Summary results for confirmatory factor analyses of versions of SHALOM.
Table 3.
Summary results for confirmatory factor analyses of versions of SHALOM.
Version of SHALOM | n | ×2/df | AIC | ECVI | IFI | NFI | TLI | PNFI | RMSEA | CFI |
Uni. | 378 | 2.70 | 574 | 1.51 | .944 | .914 | .935 | .79 | .067 | .944 |
Sec. stu. | 460 | 2.70 | 575 | 1.25 | .941 | .909 | .931 | .79 | .061 | .940 |
Generic | 409 | 3.51 | 668 | 1.64 | .931 | .906 | .920 | .78 | .078 | .931 |
perT | 231 | 2.97 | 620 | 2.70 | .899 | .855 | .882 | .74 | .093 | .898 |
impT | 139 | 1.66 | 404 | 2.93 | .940 | .882 | .930 | .74 | .069 | .939 |
gd value | | <2–3 | lowest | lowest | >.90 | >.90 | >.90 | >.50 | <.05-.08 | >.93 |
The sample in this study was split by views on the Transcendent, into personal (e.g., Allah, God,
etc) and impersonal (e.g., higher power, something there,
etc). Acceptable values for fit indices were obtained for those choosing impersonal Transcendent, but less than satisfactory values for those embracing a personal Transcendent (see
Table 3). The 39 respondents who indicated non-belief in such an entity were omitted from this study of subsets as the small number did not permit appropriate statistical calculations.
3.4. Comparing Spiritual Dissonance in Versions of SHALOM
SHALOM actually comprises two spiritual well-being measures. The Spiritual Health Measure (SHM) is the lived experience component discussed above, whereas the (Spiritual) Life-Orientation Measure (LOM) comprises the ideals for spiritual well-being held by respondents. Together they make the Spiritual Health And Life-Orientation Measure (SHALOM). The difference in score between the ideals and lived experiences reveals the level of dissonance or harmony in each domain of spiritual well-being. This double-response method has been shown to provide a better assessment of the quality of relationships reflecting each of four domains of spiritual well-being, in comparison with lived experiences only [
11]. Comparing values of fit indices in
Table 3 and
Table 4 for the university and secondary school cohorts reveals similar results, with the dissonance method generally providing better fit indices.
The fit indices for the dissonance scores on the generic form of SHALOM show acceptability for the instrument as a whole and for the impersonal Transcendent group in particular (see
Table 4). However, the dissonance values for those embracing a personal Transcendent did not yield an acceptable model, similar to the results using lived experience only (shown in
Table 3). Therefore, it can be concluded that using the word ‘God’ is more acceptable to these people than ‘Transcendent’.
Table 4.
Summary of CFA for dissonance values in versions of SHALOM.
Table 4.
Summary of CFA for dissonance values in versions of SHALOM.
Version of SHALOM | n | ×2/df | AIC | ECVI | IFI | NFI | TLI | PNFI | RMSEA | CFI |
dUni. | 378 | 2.63 | 524 | 1.39 | .907 | .859 | .892 | .74 | .066 | .906 |
dSec. stu. | 460 | 1.97 | 415 | .91 | .950 | .903 | .941 | .78 | .046 | .949 |
dGeneric | 409 | 2.81 | 553 | 1.36 | .896 | .847 | .878 | .73 | .067 | .895 |
dperT | 231 | 2.89 | 605 | 2.63 | .808 | .734 | .774 | .63 | .091 | .805 |
dimpT | 139 | 1.56 | 387 | 2.81 | .909 | .782 | .892 | .68 | .064 | .907 |
gd value | | <2-3 | lowest | lowest | >.90 | >.90 | >.90 | >.50 | <.05-.08 | >.93 |
4. Conclusions
Four of the five items in the Transcendental domain of the original Spiritual Health And Life-Orientation Measure (SHALOM) were adjusted by replacing the words ‘God, ‘Divine’ and ‘Creator’ with the word ‘Transcendent.’ Before completing the 20-items on the revised SHALOM in a web survey, respondents who had previously expressed interest in spirituality, or a likely aversion to it, were asked to indicate their description of any Transcendent influence on their lives, or if they did not believe in such an entity. Confirmatory factor analysis revealed that the modified, generic form of SHALOM showed acceptable model fit, comprising four clearly delineated domains of spiritual well-being. Of particular interest was the finding that the modified Transcendental domain of spiritual well-being holds together well statistically and provides the greatest explanation of variance in spiritual well-being overall. This modified SHALOM can be employed in future studies as a generic measure of spiritual well-being across a variety of worldviews. However, the nature of the cohort under investigation should determine which version of SHALOM is used, not the worldview of the investigators.
It should also be kept in mind that the model fit for the more generic version of SHALOM was not as good as that for the original, when it was compared with a select sample of university students and staff, and a cohort of secondary school students. This study, of the generic version of SHALOM, only had limited numbers of well-educated people in each category relating with alternative Transcendents for their spiritual well-being. Therefore further studies with more people are warranted to gain adequate sample sizes to see how well each version of SHALOM fares in general populations with varying world-views. Overall, the tentative conclusion that can be drawn from this study is that, rather than investigating relationship with a Transcendent, you can’t beat relating with God for spiritual well-being.