Brazilian Adaptation and Validation of the Religious and Spiritual Struggles (RSS) Scale—Extended and Short Version

Abstract

Studies highlight that although the spiritual/religious dimension is commonly used as a resource for coping with stress and suffering, sometimes this dimension can also be a locus of struggles. The Religious and Spiritual Struggles (RSS) Scale, composed of 26 items (extended version), is an instrument that assesses the presence of six types of spiritual struggles, categorized into three major categories: Intrapersonal, Interpersonal, and Supernatural. More recently, studies have presented a leaner version, with 14 items (short version), also suitable for evaluating religious/spiritual struggles. This study aimed to validate the RSS Scale, in its two versions, in the Brazilian cultural context and was able to attest to its quality and reliability, albeit with small variations. The Brazilian extended version presented six components (similar to the original version) explaining 68.48% of variance and 0.907 of KMO. Two items were withdrawn due to low loadings, but the psychometric qualities of both versions in two different samples (one wide sample and one with LGBTQIA+ individuals) were maintained. The reliability (Cronbach’s alpha) for each one of the six components ranged from 0.78 to 0.88 (24 items) and from 0.73 to 0.83 (14 items). The TLI and CFI were, respectively, 0.934 and 0.945 (24 items) and 0.982 and 0.988 for the short version (14 items). Furthermore, convergent validity tests indicated correlations with the Satisfaction with Life Scale. The RSS Scale (the Brazilian versions with 24 and 14 items), demonstrated reliability in the studied sample and can be applied in clinical and research contexts.

1. Introduction

According to the latest census, the population of Brazil was of 195,700,000 (Instituto Brasileiro de Geografia e Estatística—IBGE 2010). A total of 92% of the inhabitants reportedly belong to a religion or have “multiple religious belongings”. Only 8% of the population claimed to be atheist or agnostic (IBGE 2010). The population was predominantly Catholic (64.6%), followed by Pentecostals (22.2%), Spiritists (2%), and Afro-Brazilian religions (0.3%), while the remaining were distributed among Buddhism, Islam, Judaism, indigenous traditions, and different eastern religions (IBGE 2010). Although the second decade of the 21st century may have brought a significant change to the Brazilian religious map (a tendency that will probably be proven in the next national census in 2022), recent studies indicate that the Brazilian population is still “highly religious”, according to data surveyed through the Centrality of Religiosity Scale (Esperandio et al. 2019). In this cultural context, studies investigating the relationship between spirituality and health have been gaining increasing attention in recent years. Evidence comes from the increase in research groups in Spirituality and Health in graduate programs, according to a recent survey (Esperandio 2020). Brazil also falls in 13th place in the international ranking of publications in the Scopus database and “5th place in articles on medicine, psychology and nursing that include R/S, in the last five years” (Moreira-Almeida and Lucchetti 2016). A predominantly positive perspective on the impact of religion and spirituality characterizes Brazilian studies in this field. Therefore, it is urgent to investigate other aspects as well, such as the implications of negative religious/spiritual (R/S) coping and R/S struggles for health outcomes, as already pointed out by Brazilian researchers (Moreira-Almeida and Lucchetti 2016; Esperandio 2020). Such studies can contribute to a better understanding of how the R/S dimension acts or interferes in health/illness processes in different scenarios, for example, in the current context of the COVID-19 pandemic (Pargament and Exline 2022), in mental health (Ellison et al. 2013; Plante 2020; Exline et al. 2021), and in palliative care (Damen et al. 2021) and end-of-life care (Puchalski et al. 2016), among others. Thus, the adaptation and validation of the Religious and Spiritual Struggles (RSS) Scale is relevant for the advancement of studies in this field in Brazil.

What are “spiritual struggles”? What does the scale of spiritual struggles measure? Do R/S struggles have a significant relationship with life satisfaction? The present study sought to adapt and validate the RSS Scale (Exline et al. 2014) in its two versions, with 26 and 14 items, for use in the Brazilian cultural context. Our hypothesis was that the RSS Scale would manifest the same properties as those found in its initial validation study, with some necessary adaptations. We looked for a reliable version that could be used in research and in the development of specific intervention plans in the provision of spiritual care.

2. Theoretical Foundations of the Religious and Spiritual Struggles (RSS) Scale

Before presenting the theoretical foundations of the RSS Scale, it is vital to define the notions of religion, religiousness and spirituality that support this study.

Religion, according to Pargament (2013, p. 258) “is a search for significance that occurs within the context of established institutions” and also provides sense of identity, community, comfort, self-control to intimacy with others, and life transformation. Regarding the term “religiousness”, we understand it to be derived from the concept of religion. In this sense, it is related to the individual’s involvement in specific religious activities, such as prayers, attendance at worship services, masses, and the most varied types of encounters promoted by some form of established religion. In general, people who consider themselves to be religious assume certain ethical-moral beliefs, practices, and values linked to an established religion. Spirituality, in turn, is a more complex notion. We follow the international consensus that defines spirituality as “a dynamic and intrinsic aspect of humanity through which persons seek ultimate meaning, purpose, and transcendence, and experience relationship to self, family, others, community, society, nature, and the significant or sacred. Spirituality is expressed through beliefs, values, traditions, and practices” (Puchalski et al. 2014, p. 644).

The authors of the original RSS characterize R/S struggles as tensions, conflicts, confrontations, and crises related to what the person considers sacred (Exline 2013; Pargament et al. 2005) and that “occur when some aspect of R/S belief, practice or experience becomes a focus of negative thoughts or emotions, concern or conflict” (Exline et al. 2014, p. 208). The authors note that facing spiritual struggles is not an uncommon experience. However, in general, people who experience them may feel uncomfortable admitting such experiences, as they are often accompanied by guilt, shame, and even fear of being criticized by others (Exline and Grubbs 2011).

R/S struggles go beyond expressions of negative affect or manifestation of personality traits (Piedmont et al. 2009). They contribute significantly, as an independent factor, to depression, anxiety, stress symptoms, and well-being (Fox and Piedmont 2020; Wilt et al. 2017). Assessing R/S struggles is essential, allowing the health care team to include this dimension in patients’ caring plans. Validating the RSS among different cultures is crucial to generalize empirical findings and deepen the understanding of the role of struggles in individuals and groups.

The six types of R/S struggles (

Table 1. Types of religious/spiritual (R/S) struggles (Exline et al. 2014).

Categories	Types of R/S Struggles	Sample Items
Intrapersonal (ME; DO; MO)	Meaning (ME) Refers to questions about a deeper meaning in life	ME1 Had concerns about whether there is any ultimate purpose to life or existence ME2 Felt as though my life had no deeper meaning ME3 Questioned whether life really matters ME4 Questioned whether my life will really make any difference in the world
	Doubts (DO) Refers to feeling troubled by doubts or questions about R/S beliefs	DO1 Struggled to figure out what I really believe about religion/spirituality DO2 Felt troubled by doubts or questions about religion or spirituality DO3 Felt confused about my religious/spiritual beliefs DO4 Worried about whether my beliefs about religion/spirituality were correct
	Moral (MO) Conflicts about following moral principles or guilt in response to perceived transgressions	MO1 Felt guilty for not living up to my moral standards MO2 Worried that my actions were morally or spiritually wrong MO3 Wrestled with attempts to follow my moral principles MO4 Felt torn between what I wanted and what I knew was morally right
Interpersonal (I)	Interpersonal (I) Refers to negative experiences with religious people or institutions or conflict with others around religious issues	I1 Had conflicts with other people about religious/spiritual matters I2 Felt rejected or misunderstood by religious/spiritual people I3 Felt as though others were looking down on me because of my religious/spiritual beliefs I4 Felt angry at organized religion I5 Felt hurt, mistreated, or offended by religious/spiritual people
Supernatural (DI; DE)	Divine (DI) Negative emotion or conflict centered on beliefs about a deity or a perceived relationship with a deity”	DI1 Felt angry at God DI2 Questioned God’s love for me DI3 Felt as though God had abandoned me DI4 Felt as though God was punishing me DI5 Felt as though God had let me down
	Demonic (DE) Concern that the devil or evil spirits are attacking an individual or causing negative events”.	DE1 Felt attacked by the devil or by evil spirits DE2 Felt as though the devil (or an evil spirit) was trying to turn me away from what was good DE3 Worried that the problems I was facing were the work of the devil or evil spirits DE4 Felt tormented by the devil or evil spirits

) may fall within three general categories: (1) intrapersonal (refers to internal and intrapsychic crises involving doubts about faith or about moral and meaning issues); (2) interpersonal (expressed as tensions around R/S issues that occur within the family, between friends, or involving religious leaders or people linked to a community of faith); and (3) supernatural (this category gathers the R/S struggles that are expressed through negative thoughts and emotions, such as anger, anxiety, fear, feelings of abandonment, disappointment, and punishment, related to the idea of god or demons. Such emotions are experienced as arising from a dimension external to the individual). The response options of the 26 items are recorded using a Likert scale of five points (1 = not at all; 5 = a great deal).

In Brazil, three measures have commonly assessed multiple domains of R/S struggles (negative R/S coping): the full Religious Coping Scale (RCOPE), with 87 items (Panzini 2004); the Brief RCOPE, with 49 items (Panzini and Bandeira 2005); and the Spiritual/Religious Coping Scale—SRCOPE 14 (Esperandio et al. 2018), which has two subscales, positive and negative religious coping, each with seven items. Although SRCOPE 14 is a practical and useful measure for application in clinical settings and research, it assesses five items of divine struggles, only one demonic item, one interpersonal item, and no intrapersonal items (Exline et al. 2014, p. 209). Thus, as highlighted by Exline et al. (2014, p. 209), “a need remains for a relatively brief measure that can assess supernatural, interpersonal, and intrapersonal struggles”.

Use of RSS in the Global Context

Studies indicate that the RSS can be applied in different contexts and with individuals of different R/S beliefs. A search on the PsyInfo, PubMed, Web of Science and Scopus databases with the terms “Religious and Spiritual Struggles Scale”, after eliminating the duplicated references, found 67 studies that used the scale or referred to it. As for the studied population, most studies still focus on samples of university students and North American adults (Stauner et al. 2016; Grubbs et al. 2016; Exline et al. 2014; Breuninger et al. 2019; Trevino et al. 2019). However, studies were also carried out in Israel (Abu-Raiya et al. 2015a, 2016), Indonesia (Aditya et al. 2019), Poland (Zarzycka et al. 2018); Colombia (Currier et al. 2019) and the Czech Republic (Janů et al. 2018). Christianity has been the most common of the religious affiliations among participants in these studies (Stauner et al. 2016; Grubbs et al. 2016; Exline et al. 2014), and other religions and samples are less studied. For example, only a few studies have been conducted with Jews (Abu-Raiya et al. 2015a), Muslims (Abu-Raiya et al. 2016), and atheists (Sedlar et al. 2018). We found only four studies, beyond the initial validation study of the RSS (Exline et al. 2014) that performed exploratory and confirmatory analyzes (Zarzycka et al. 2018; Abu-Raiya et al. 2015a, 2016; Janů et al. 2018).

The Polish (Zarzycka et al. 2018) and the Iraelian (Jews sample) (Abu-Raiya et al. 2016) validation of the RSS revealed six subscales in the exploratory analysis. In the confirmatory analysis, the RSS showed good indices of CFI, TCLI, RMSEA, and SRMR in both studies. Thus, in these samples, the RSS showed psychometric properties similar to those in the initial validation study. In the other Israeli study, with Muslims (Abu-Raiya et al. 2015a), the psychometric results of the scale were slightly different and a 5-factor model emerged. Three of these factors were identical to the initial validation study (Interpersonal, Moral, and Meaning of Life). Items related to Doubt and four items related to the Divine were grouped into a single factor, and the Demonic factor included one item (inverse) that was originally in the Divine factor (Abu-Raiya et al. 2015a). Thus, in the same country (Israel), depending on the religious group, the RSS presented different numbers of factors: six for Jews (Abu-Raiya et al. 2016) and five for Muslims (Abu-Raiya et al. 2015a).

The Czech study showed different results. The sample was representative of residents of the Czech Republic, and consisted of 1000 participants aged between 15 to 90 years. The exploratory analysis with six factors grouped together items related to the divine (five items) and demonic (four items) domains; however, four of these nine items had factor loadings between 0.2 and 0.5. The four items of the Meaning domain formed a factor, but one of the items had a factor loading of 0.29. In the Moral domain, the four items formed a factor, all with factor loadings greater than 0.5. On the other hand, the nine items of the Doubt and Interpersonal domains came together to create a factor, but four item loadings were between 0.17 and 0.49. Thus, only the Moral domain showed a similar grouping to other studies using the RSS. Based on this arrangement of items and their intercorrelations, instead of doubling the subscales the authors proposed a model with only three of the original six subscales. Consequently, they reduced the RSS to items referring to the Divine, Interpersonal, and Meaning domains. This three-factor model showed a SRMR of 0.035, CFI of 0.998, TLI of 0.998, and RMSEA of 0.037 (Janů et al. 2018).

The RSS was also used in Colombia and Indonesia, but without further validation studies. The Colombian study (Currier et al. 2019) intentionally suppressed the Demonic factor, resulting in a 22-item and 5-factor scale. It only assessed internal consistencies, which ranged between 0.8 and 0.92. The authors who used the scale in Indonesia (Aditya et al. 2019) reported an overall reliability for the 26 items of 0.91 and used a score with the mean of items.

Two recent studies, one conducted by Damen et al. (2021) and another by Exline et al. (2021), used a concise version of the RSS with 14 items (RSS-14). The study by Damen et al. (2021), carried out in the United States, had a sample of 331 palliative care patients. They presented only limited psychometric information on the 14 items, namely internal consistency estimates through Cronbach’s alpha (0.79). The study by Exline et al. (2021) had 315 participants, who were transgender and gender-nonconforming adults from North America and Western Europe, and although it does not present exploratory and confirmatory analysis of the items, the study makes reference to the short version, which has undergone initial structural analyses and validation (RSS-14; Exline et al. n.d.). Reduced versions of scales tend to have a higher quality of responses as they reduce fatigue and promote greater attention from the participants. However, initial structural, reliability, and validity tests of the RSS-14 have shown promising results in a large sample of U.S. adults (Exline et al. n.d.). To maintain the consistency and clarity of the two versions of the scale, the term RSS is used in reference to the initial validation with 26 items (Exline et al. 2014) and RSS-14 to refer to the short version (Exline et al. n.d.).

3. Method

The process of translation, adaptation and validation of the RSS for its use in the Brazilian context went through a number of phases following the protocol established in the literature (Beaton et al. 2000). Before the initial translation, authorization from the authors of the original scale was sought to carry out the study. The final Portuguese version was part of the set of instruments that assessed spirituality and religiousness in the study entitled “(In)visible Resources: Evaluation of spiritual/religious coping and spiritual struggles in health”.

3.1. Translation and Adaptation

The translation and adaptation were done in four stages: Initial Translation, Expert Assessment, Pilot Study, and Reverse Translation.

Stage 1—The initial translation from English to Portuguese (Brazil) was performed by a certified translator, maintaining its original format: a general instruction and the presentation of 26 items with response options on a 5-point Likert scale (1 = not at all/not applicable; 5 = very much).

Stage 2—Evaluation of the Portuguese version by experts. For the analysis of the Portuguese version, a specific evaluation form was prepared in order to list the suggestions for improving the translation. This form was sent to four specialists as an attachment to the invitation in which the protocol and goal of the study were explained. Experts, familiar with the topic and the scale adaptation processes, analyzed the clarity, comprehensibility, and adequacy of the original meanings of the scale in English and Portuguese.

Stage 3—Focus Group/Pilot Study. After summarizing the translations and incorporating the experts’ observations, a first Portuguese version of the scale was applied to 42 Theology students in a pilot study. They served as a focus group to discuss the semantic clarity of the scale and assess whether religious expressions could be understood by the general population, regardless of their religion and beliefs. In statements 7 (“felt attacked by the devil or by the evil spirits”) and 11 (“felt as though the devil (or an evil spirit) was trying to turn me away from what was good”), the word “devil” was replaced by the expression “evil forces” and “evil spirit”. The devil is part of Brazilian religious beliefs. However, participants that evaluated the scale (focus group/pilot study) felt more comfortable with the terms “evil forces” and “evil spirits”. From their perspective, the general population would also better understand and feel comfortable with the terms “evil forces” and “evil spirits” than the word “devil”. Thus, the following modifications were made to the items on this subscale:

-

Item 7 of the original Scale: “felt attacked by the devil or by evil spirits”.

-

Portuguese translation suggested by the focus group: “Senti-me atacado por ‘forças do mal’ ou por espíritos malignos”.

-

Reverse translation: “I felt attacked by ‘evil forces’ or evil spirits”.

-

Item 11 of the original Scale: “felt as though the devil (or an evil spirit) was trying to turn me away from what was good”.

-

Portuguese translation suggested by the focus group: “Senti como se um ‘espírito do mal’ estivesse tentando me afastar do caminho do bem”.

-

Reverse translation: “I felt as if ‘an evil spirit’ was trying to tear me from the good walk”.

The other items remained with meanings equivalent to the originals, resulting in a final version of the scale in Portuguese.

Stage 4—Reverse translation. The Portuguese version was sent to three North Americans residing in Brazil for reverse translation. These translations were sent to the author of the original scale, who approved the final Portuguese version.

3.2. Validation

The data collection from the general population was done through the wide distribution of an electronic link on social media (e-mail addresses and Facebook contacts of the researchers and students involved in the project). The form was developed on the Qualtrics Platform, where the data continue to be hosted. Anonymity and one time participation was guaranteed. In addition to having access to a computer with internet, participants had to be over 18 years of age. All of the participants gave their informed consent for inclusion before participating in the study, which was conducted in accordance with the declaration of Helsinki, and the protocol was approved by the PUCPR Research Ethics Committee (Process 1.354.361). The data collection among the LGBTQIA+ population was carried out through invitation in groups that gather activists, gender and sexuality researchers, sexual minorities, and NGOs for the protection and defense of the rights of these populations. The research was approved by the Research Ethics Committee (Process 4.122.312). The validation study included exploratory analysis, confirmatory analysis, and convergent validity, performed with a sample of the general population, and confirmatory analysis and convergent validity with the LGBTQIA+ sample.

The sample was randomly divided for Exploratory Factor Analysis (EFA) (Sample EFA—composed of 293 participants, 49.1% of total valid sample) and for Confirmatory Factor Analysis (CFA) (Sample CFA 1—with 304 participants, 50.9% of total valid sample). Each one of the Samples (EFA and CFA 1) attended most recommendations in the literature, and had the minimum of 200 participants for EFA (Guilford 1954; MacCallum et al. 1999; Hair et al. 2010). After EFA analysis (Sample EFA), we conducted a CFA analysis (Sample CFA 1) to verify whether the overall number of observed variables could be reduced into latent dimensions, based on their commonalities. According to McArdle (1996), this is done to reduce measurement error and add a level of statistical precision that is useful for scale validation. In CFA, we followed the recommendation of Hair et al. (2010), testing the model for its convergent (dimension of common variance) and discriminant validity (assumes that one dimension is different from other ones). CFA was also applied to assess the validity of a short version of the RSS for these Brazilian samples.

To test convergent validity, CFA was applied to the LGBTQIA+ sample (Sample CFA 2). We also tested convergent validity with the Satisfaction with Life Scale (Sample EFA + CFA 1).

3.3. Measure for Correlating Tests: Satisfaction with Life Scale (SWLS)

All domains of RSS have been negatively correlated with life satisfaction (Abu-Raiya et al. 2015b). Life satisfaction and RSS scores were also intercorrelated in an atheist sample (Sedlar et al. 2018). The SWLS is a 5-item scale designed to measure global cognitive judgments of one’s life satisfaction (Diener et al. 1985). SWLS is widely used and validated in Brazil across diverse samples and different cultural contexts (Gouveia et al. 2009; Silva et al. 2021; Zanon et al. 2014). It is not a measure of either positive or negative affect. Participants indicate how much they agree or disagree with each of the five items using a 7-point scale that ranges from 7 strongly agree to 1 strongly disagree. SWLS is a validated, consistent, and user-friendly instrument. Moreover, life satisfaction is a common outcome related to RSS scores. Thus, we selected this measure to provide an initial test of convergent validity of the RSS.

4. Results

4.1. Sample: Demographic Characteristics

A total of 796 individuals from several different states in Brazil started to answer an on-line Qualtrics questionnaire sent by e-mail and through social media. In total, 597 individuals completed it (age MD = 37.93, SD = 13.75). Most were women (65.5%) and 91.5% had a higher education (24.8% were college students, 29.5% had college degrees, and 37.2% had higher degrees—MBA, PhD, etc.), 7.0% finished high school, and 1.5% did not complete school. Most of the sample had a general income below eight minimum wages (at the time it was equivalent to USD 264.00/month), so 18.3% were paid less than one minimum wage, 29.0% between one and three, 28.8% less than eight, and only 24.0% had an income above it—similar to the Brazilian socioeconomic pyramid. Most of the participants were married (40.8%) or single (41.4%), 10.5% with stable union, 5.9% were separated/divorced, and 1.3% widowers. The sample included participants from different religions, Catholics represented the largest group (33.7%), Evangelicals were the second largest group (24.5%), and Spiritists the third (12.1%). Approximately 13.6% affirmed belief in God but have no religion. Furthermore, smaller percentages of other religions were reported: African-Brazilian, Pentecostal, Buddhism, and combinations of different religions.

Regarding these socio-demographic descriptions, we found no differences between the samples (EFA vs. CFA 1) or RSS scores (p > 0.100). Additionally, women and men demonstrated no difference in their RSS score (M_women = 1.87, SD = 0.62/M_men = 1.88, SD = 0.65/t = 0.129, p = 0.897).

4.2. Exploratory Analysis—Sample EFA

Results showed that the sample was adequate for a factor analysis (KMO = 0.907, Bartlett’s sphericity −χ² = 4335.561, p < 0.001, df = 325). The first attempt used the K1 method proposed by Kaiser (1960) and eigenvalues analysis (Cattell 1966) presented a five-factor solution with 65.20% of total variance explained. Still, it showed that six factors would explain 68.48% of the total variance.

Table 2. Rotated component matrix of the six-factor solution.

Dimension		1 (14.6% var.)	2 (13.5% var.)	3 (11.6% var.)	4 (10.2% var.)	5 (9.9% var.)	6 (8.6% var.)
Divine	DI1	0.772
	DI2	0.802
	DI3	0.826
	DI4	0.695
	DI5	0.795
Demonic	DE1		0.849
	DE2		0.863
	DE3		0.792
	DE4		0.773
Interpersonal	I1			0.676
	I2			0.773
	I3			0.438
	I4			0.719
	I5			0.798
Moral	MO1						0.633
	MO2						0.731
	MO3						0.712
	MO4						0.506
Meaning	ME1				0.646	0.431
	ME2	0.405			0.637
	ME3				0.739
	ME4				0.716
Doubt	DO1					0.528
	DO2					0.649
	DO3					0.631
	DO4					0.736

presents the six-factor solution with a Varimax Rotation, suppressing loadings lower than 0.4. Due to these results, the CFA was tested with all 26 items of the 6 dimensions proposed by Exline et al. (2014).

4.3. Confirmatory Analysis

The first model tested in AMOS v.20 for SPSS v.20 (Sample CFA 1) resulted in some acceptable fit indexes (χ²/df = 2.092, AGFI = 0.836, RMR = 0.070, SRMR = 0.062, RMSEA = 0.060, GFI = 0.871—recommended by Saris and Stronkhorst 1984; Hair et al. 2010), but not all of them (some below acceptable fit indexes: TLI = 0.901 ≥ 0.92 for acceptance and CFI = 0.916 ≥ 0.92 for acceptance). The Composite Reliability (CR) of the six dimensions was over 0.7 (showing high reliabilities—Hair et al. 2010), but two dimensions had AVE (Average Variance Extracted) below acceptable (Divine AVE = 0.47/Moral AVE = 0.44) (AVE must be ≥ 0.50—Hair et al. 2010).

We analyzed standard coefficients estimates, and they demonstrated that two items (DIVINE—DI4 “Felt as though God was punishing me” coeff = 0.453, and MORAL MO3 “Wrestled with attempts to follow my moral principles” coeff = 0.356) had low coefficients when compared to all others (>0.600) and high covariance (within dimensions and, also other dimensions’ items).

Therefore, we tested a second model with the original RSS six dimensions (Demonic, Interpersonal, Meaning, Divine, Moral, and Doubt) but without these two items (DI4 and MO3). The results presented overall good fit indexes (χ²/df = 1.801, TLI = 0.934, CFI = 0.945, GFI = 0.900, AGFI = 0.867, RMR = 0.063, SRMR = 0.054, RMSEA = 0.051). The Composite Reliability (CR) for all six dimensions was over 0.7 (showing high reliabilities—Hair et al. 2010), and all six dimensions had AVE (Average Variance Extracted) (

Table 3. Results for RSS with six dimensions: standardized regression weights, AVE, and CR.

Item	Dimension	St. Reg. Weights		CR		AVE
		24 Items	14 Items	24 Items	14 Items	24 Items	14 Items
DI5	Divine	0.686	0.777	0.82	0.76	0.53	0.52
DI3		0.752	0.654
DI2		0.851
DI1		0.609	0.726
DE4	Demonic	0.780		0.88	0.83	0.64	0.71
DE3		0.757	0.866
DE2		0.839
DE1		0.824	0.819
I5	Interpersonal	0.635		0.84	0.77	0.50	0.53
I4		0.776	0.760
I3		0.696
I2		0.730	0.727
I1		0.706	0.690
MO4	Moral	0.738		0.78	0.73	0.55	0.57
MO2		0.801	0.844
MO1		0.678	0.659
ME4	Meaning	0.653		0.80	0.75	0.51	0.60
ME3		0.760	0.697
ME2		0.767	0.847
ME1		0.664
DO4	Doubt	0.761		0.84	0.77	0.57	0.62
DO3		0.795	0.783
DO2		0.814	0.794
DO1		0.635

CR = Composite Reliability; AVE = Average Variance Extracted.

) above acceptable, with confirmed discriminant validity—all AVE were higher than the corresponding squared correlation coefficient between factors (Fornell and Larcker 1981) (

Table 4. Results for RSS with six dimensions: correlations, squared correlations *, AVE, and CR.

Dimension	Divine	Demonic	Interp.	Moral	Meaning	Doubt
Divine	1.000	0.114	0.120	0.114	0.466	0.413
Demonic	0.338	1.000	0.054	0.158	0.069	0.097
Interpersonal	0.347	0.232	1.000	0.056	0.201	0.326
Moral	0.338	0.398	0.236	1.000	0.192	0.379
Meaning	0.681	0.262	0.448	0.438	1.000	0.471
Doubt	0.643	0.312	0.571	0.616	0.686	1.000
AVE	0.53	0.64	0.50	0.55	0.51	0.57
CR	0.82	0.88	0.84	0.78	0.80	0.84

* Correlations below the diagonal and squared correlations are above the diagonal. CR = Composite Reliability; AVE = Average Variance Extracted.

). Additionally, all dimensions’ correlations were below 0.90, rejecting multicollinearity (Hair et al. 2010), and the general internal consistency for RSS was acceptable (0.801) (Sample EFA = 0.831/Total Sample = 0.764). A comparison with Exline et al. (2014) demonstrated that, besides the exclusion of DI4 (also the lowest factor loading of the dimension in the original scale) and MO3, factor loadings were quite similar, although a little lower. There were a few exceptions: the factor loadings on DI1, MO1, and DO1 were lower than the initial validation study, 0.80, 0.81, and 0.85 (Exline et al. 2014), respectively, but overall they maintained similar psychometric results.

4.4. Short Version—Confirmatory Analysis

We also conducted a CFA analysis on a short version of the RSS, using the RSS-14 and its subdomains as the initial basis (Damen et al. 2021; Exline et al. n.d.), but replacing MO3 with MO2 and DI4 with DI5, due to low standardized regression weights and high covariance. These replacements resulted in acceptable indexes (χ²/df = 1.281, TLI = 0.982, CFI = 0.988, GFI = 0.963, AGFI = 0.935, RMR = 0.034, SRMR = 0.032, RMSEA = 0.030) (AVE: Divine [DI5;DI3;DI1] = 0.52; Demonic [DE3;DE1] = 0.71; Interpersonal [I4;I2;I1] = 0.53 Moral [MO2;MO1] = 0.57; Meaning [ME3;ME2] = 0.60; Doubt [DO3;DO2] = 0.62) (CR: Divine = 0.76; Demonic = 0.83; Interpersonal = 0.77 Moral = 0.73; Meaning = 0.75; Doubt = 0.77) and discriminant validity (dimensions’ correlations were below 0.90, and all AVEs were higher than the corresponding squared correlations—

Table 5. RSS six dimensions correlations and square correlations.

			Sample CFA 1		Sample CFA 2 (LGBT+)
			24 Items	14 Items	24 Items	14 Items
			r	r	r	r
DI	<-->	DE	0.338	0.358	0.135	0.292
DI	<-->	I	0.347	0.469	0.356	0.437
DI	<-->	MO	0.338	0.223	0.463	0.429
DI	<-->	ME	0.681	0.702	0.467	0.550
DI	<-->	DO	0.643	0.619	0.527	0.568
DE	<-->	I	0.232	0.220	0.098	0.134
DE	<-->	MO	0.398	0.364	0.402	0.444
DE	<-->	ME	0.262	0.324	0.126	0.201
DE	<-->	DO	0.312	0.257	0.236	0.304
I	<-->	MO	0.236	0.140	0.462	0.414
I	<-->	ME	0.448	0.417	0.424	0.406
I	<-->	DO	0.571	0.644	0.698	0.754
MO	<-->	ME	0.438	0.340	0.450	0.335
MO	<-->	DO	0.616	0.542	0.733	0.744
ME	<-->	DO	0.686	0.662	0.532	0.521

, when comparing versions and samples results), suggesting good preliminary results for the RSS-14 version for Brazil.

Using the full sample (Sample EFA + CFA 1, N = 597), we investigated the convergence of both versions of the RSS with the Satisfaction with Life Scale (Diener et al. 1985) (Cronbach’s Alpha = 0.839) (

Table 6. RSS six dimensions: RSS correlations with Satisfaction with Life.

		Satisfaction with Life
Divine	24 items	−0.378 **
	14 items	−0.367 **
Demonic	24 items	−0.248 **
	14 items	−0.248 **
Interpersonal	24 items	−0.261 **
	14 items	−0.247 **
Meaning	24 items	−0.445 **
	14 items	−0.467 **
Doubt	24 items	−0.257 **
	14 items	−0.279 **
Moral	24 items	−0.245 **
	14 items	−0.192 **
RSS	24 items	−0.429 **
	14 items	−0.409 **

** p < 0.001.

): the 24-item RSS (Cronbach’s Alphas: Divine = 0.847, Demonic = 0.878, Interpersonal = 0.823, Moral = 0.793, Meaning = 0.815, Doubt = 0.805) and the RSS-14 version (Cronbach’s Alphas: Divine = 0.789, Demonic = 0.793, Interpersonal = 0.727, Moral = 0.748, Meaning = 0.769, Doubt = 0.765). The results demonstrated a similar pattern on both RSS scale versions, such that when individuals had lower RSS scores they presented higher satisfaction with life. Interestingly, the dimensions with higher correlations were Meaning and Divine.

4.5. Second Sample CFA Analysis

We examined the consistency of findings by testing the RSS versions (24 items and short 14 items) on a sample that was collected about two years later (LGBTQIA+: Sample CFA 2, n = 298). The affective-sexual orientations represented were: homosexuals (61%); bisexuals (28.2%); pansexual (6.5%); asexual (1.9%); heterosexuals (1.6%); and questioning (0.6%). As for marital status, 76% were single; 21.1% were married or in stable unions; and 2.9% were separated or divorced. Almost half of the sample (41.2%) had less than a complete higher education; 29.9% had postgraduate studies; 20.8% had completed higher education (mainly courses in the area of Health: Nursing, Medicine and Psychology); 6.5% had finished high school; and 1.6% did not finish high school.

The results revealed overall good fit indexes for the 24 items (χ²/df = 2.262, TLI = 0.906, CFI = 0.921, GFI = 0.873, AGFI = 0.835, RMR = 0.092, SRMR = 0.065, RMSEA = 0.062/AVE: Divine = 0.62; Demonic = 0.60; Interpersonal = 0.53 Moral = 0.56; Meaning = 0.60; Doubt = 0.55/CR: Divine = 0.86; Demonic = 0.86; Interpersonal = 0.84 Moral = 0.79; Meaning = 0.86; Doubt = 0.83). The same six-factor solution was valid for the 14 items (χ²/df = 1.923, TLI = 0.948, CFI = 0.966, GFI = 0.950, AGFI = 0.912, RMR = 0.052, SRMR = 0.041, RMSEA = 0.056/AVE: Divine = 0.55; Demonic = 0.60; Interpersonal = 0.58 Moral = 0.58; Meaning = 0.77; Doubt = 0.63/CR: Divine = 0.79; Demonic = 0.74; Interpersonal = 0.8 Moral = 0.73; Meaning = 0.87; Doubt = 0.77). Furthermore, both versions yielded evidence of discriminant validity (dimensions’ correlations were below 0.90, and all AVEs were higher than the corresponding squared correlations—also Table 5).

5. Discussion

Findings from exploratory and confirmatory analyses of the Brazilian version of the RSS (26-item version) were comparable to the original scale developed by Exline et al. (2014). Two items out of twenty-six weakened the parameters in the confirmatory analysis, suggesting a version composed of twenty-four items for the Brazilian RSS version. The Brazilian short version of the RSS (14 items) presented the same number of items with only two replacements (MO3 with MO2, DI4 with DI5). The Brazilian version of the RSS-14 presented slightly better fit indexes than the extended (24 items) version of the RSS, suggesting greater applicability of this shorter scale as it would be less time-consuming to collect data from participants. We tested the exploratory and confirmatory parameters in two different populations, and the results remained acceptable. In studies that use several scales, the time needed to answer questionnaires and scales is a factor that reduces the quality and quantity of participation. Thus, the 14-item scale can be a potentially valuable tool, as long as it maintains levels of validity and reliability that are similar to the extended version. The 14-item scale can be analyzed in terms of both its overall score (average of 14 items) and scores on the six dimensions (Sandage et al. 2022; Damen et al. 2021; Exline et al. 2021, n.d.). This validation study presented good results when using the 14-item scale as an overall score and in domains comparable to other studies in other cultures. The domains remained the same as those in the original scale, with two or three items in each of them.

The use of six dimensions for the scale showed excellent parameters, indicating that the dimensionality of the Brazilian RSS (24-item) version is congruent with that of the original work of Exline et al. (2014). The five-dimension option (EFA), which merges the Moral and Doubt dimensions, also achieved satisfactory results; however, the use of six dimensions was similar to the five dimensions, improved the explained variance of the measure, and was closer to the original content and dimensionality proposed by Exline et al. (2014). Some of the RSS validity studies in other cultures yielded at least somewhat different dimensionality; for example, the adaptation for Muslims in Israel, which resulted in five dimensions and merged the Doubt and Divine domains (Abu-Raiya et al. 2015a), and there is the Czech adaption, in which a solution of three domains was found (Janů et al. 2018). However, most of the adaptation studies supported the dimensionality in the six domains that were found in the original scale (Zarzycka et al. 2018; Abu-Raiya et al. 2016; Exline et al. 2014). The United States, Poland, and Brazil seem to share a similar scale grouping structure. In Israel, there appears to be some divergence, depending on the population studied, but no significant changes in scale. On the other hand, the results of the Czech Republic study show an organization that is quite different from the original. Thus, although the R/S struggle domains may group together differently in some contexts, the six-domain solution seems to be the most consistent one across different cultures.

During the exploratory analysis, two items showed low standard coefficients estimates: MO3—“Wrestled with attempts to follow my moral principles” and DI4—“Felt as though God was punishing me”. Excluding these items improved the CFA parameters considerably. These results suggest that “punishment by God” may not be closely related to other types of Divine struggle (abandonment, anger at God) in the Brazilian context. That, in turn, suggests we might need to measure punishment by God more fully in future studies. These findings may indicate a change in the way religiousness has been experienced in the Brazilian culture, e.g., moving away from more traditional forms of religious expression (where guilt occupied a prominent place allied with the idea of divine punishment) toward shame and other aspects, such as happiness, prosperity, health, and wealth (where evil spirits are responsible for diseases, poverty, and morally incorrect or harmful behavior) (Esperandio 2007). These inferences should be investigated in future studies.

The convergence of results was assessed between populations by applying the two forms of the scale in a group that was different from the one initially selected for the study. An application with the LGBTQIA+ population was recently performed by Exline et al. (2021). Here, confirmatory analyzes for the extended (24 items) and short (14 items) Brazilian versions presented very similar indices to the population used in the validation of our study. Furthermore, the results of the two versions were comparable and showed excellent fit indices. Life satisfaction was also correlated to Brazilian version of the RSS and RSS-14 among the domains of the scale and also to the global score of all items.

R/S struggles appear to be present in different cultures and contexts. The associations of these struggles with life satisfaction and other fundamental health outcomes such as clinical impairment (Fox and Piedmont 2020), symptom burden, and poor quality of life (Damen et al. 2021) reinforce the importance of attending to R/S struggles not only in empirical and research studies, but also in clinical and spiritual care practice. The role of R/S struggles, and the validity of its multiple domains must be better studied. However, the persistency of a multifaceted structure of R/S struggle domains emerging from research in different cultures underscores the rich and varied nature of struggles in the R/S realm. Moreover, it is possible that the domains can play different roles in psychological functioning. Even in an atheist sample R/S struggles were associated with depression, anxiety, lower life satisfaction, and a lower sense of meaning in life (Sedlar et al. 2018). However, in contrast to other samples, in which most all of domains are associated with distress and well-being, Sedlar et al. (2018) found that, among atheists, only Interpersonal, Moral, and Ultimate Meaning struggles were associated with these life outcomes.

6. Conclusions

The RSS has shown a common structure among different samples; however, some adaptations in the number of items in the subscales may be necessary. The RSS is a validated measurement tool and deserves further study to deepen our understanding of R/S struggles among diverse cultures and contexts and allow for better care and health outcomes. This RSS validation study corroborates the use of the scale with the Brazilian population in the extended version. However, we suggest the exclusion of two items from the original 26-item scale, resulting in a Brazilian version with 24 items. Both the 24-item and 14-item versions presented good fit indexes and are recommended, but, when time and simplicity for collection is necessary, the short version may be a good alternative.