1. Introduction
The phenomenon of religiosity constitutes a complex and multifaceted reality that is investigated on different levels of scientific reflection: theological, philosophical, psychological, and sociological levels. This is due to the fact that religiosity comprises such forms as religious beliefs, personal and communal worship, religious experience, or emotions, which manifest themselves in different spheres of personal and social life (
Głaz 2021a;
Koenig et al. 2015). The aforementioned scientific disciplines are based on their own assumptions and use specific methods to describe and determine the phenomenon of religiosity. The recent polls conducted by the Gallup Poll Social Series (GPSS) found that around 81% of Americans believe in God, and 42% are convinced that God hears their prayers (
The Gallup Poll Social Series 2022). In Poland, 81% of the population regard themselves as believers or firm believers, 46% attend a mass or a religious service at least once a week, and around 70% pray daily or at least once a week (
Statistics Poland 2019). The development of new objective and constructive measures of religious practice is, therefore, integral to a psychological understanding of the role played by religion in contemporary life (
Hill 2013). Despite the growing literature on religious practices, few reported psychological measures refer specifically to Catholic religious practices.
The aim of the present study is to develop a reliable and valid questionnaire that would assess Catholic religious practices conceptualized within the dogmatic and liturgical norms of the Catholic Church. Within this objective, we will assess the psychometric qualities of the questionnaire: the factor structure through exploratory factor analysis, the association between manifest variables and latent factors through confirmatory factor analysis (CFA), internal reliability, and convergent validity.
3. Study 1
3.1. Development of Original Item Pool
The aim of Study 1 was to generate a pool of statements that were related to a preliminary conceptual definition of religious practices in the context of liturgy and folk piety. The identification process was based on a review of the literature (deductive method) and responses from individuals (inductive method) as such an approach is regarded as best practice in item development (
Boateng et al. 2018). The
Directory on Popular Piety and Liturgy was one of the main sources of inspiration for creating items. In addition, we included people’s views on their practice of faith. Due to the specificity of the planned questionnaire, we did not use other statements from the existing scales. In this sense, all items were originally created by the research team. The main criterion adopted in the creation of items was a reference to the manifestations of official religiosity (e.g., liturgy) and folk piety (e.g., cultural elements), understood as complementary forms of expressing faith.
Following the guidelines that the initial item pool should contain at least twice as many items as the intended questionnaire (
Morey 2013), the authors of this study created 40 stimulus items. They also tried to keep the content of the items simple and non-ambivalent (
Boateng et al. 2018), and included reverse-scored items to reduce some potential acquiescence bias (
Field and Miles 2010). Four experts in the field of theology and psychology examined the congruency of items. They evaluated all items considering the compliance of the content of each statement with the criterion referring to official religiosity and popular piety. They also paid a special attention to the character of Catholic traditions. Based on their evaluations, the pool was reduced to 15 items (
Table 1).
All items were assessed on a 7-point Likert scale with 1 = strongly disagree; 2 = disagree; 3 = partially disagree; 4 = neither agree nor disagree; 5 = partially agree; 6 = agree; and 7 = strongly agree. In terms of content, the statements reflected the two-dimensional structure of the CRPQ, referring to folk practices and official religiosity.
3.2. Participants
Sample 1 comprised 219 participants between 18 and 87 (M = 26.92; SD = 11.39) and included 146 women (67%) and 73 men (33%). In terms of place of residence, 25% of respondents indicated the countryside, 17%—a city up with to 25,000 inhabitants, 13%—a city between 25,000 and 100,000 inhabitants, and 25%—a city above 100,000 inhabitants. When asked to what extent they believe that God exists, the average obtained was M = 7.10 (SD = 3.34), on a scale from 0 to 10. The connection with the Catholic Church was at the level of M = 6.00 (SD = 3.72). When it regarded participation in the Holy Mass, most people (43%) indicated the Sunday Mass, followed by those who go to Mass only on major liturgical feasts (22%), every day (7%), and once every few weeks (6%). Exactly 22% of the respondents declared that they never attended Holy Mass. When asked whether they belong to any group related to the Catholic Church, the respondents who answered positively, indicated the following communities: University Chaplaincy Centre, Light-Life Movement, Liturgical Service of the Altar, Living Rosary, Priesthood, Religious Congregation, Evangelizing Community, Catholic School, Catholic Youth Association, Parish Pastoral Care, Diocesan Diaconia, Caritas Polska, Vincentian Marian Youth, Bible Circle, and Church Choir.
Participants were recruited through purposive sampling, using different internet-based networks and email correspondence that was addressed to individuals who declared their affiliation with the Catholic Church.
3.3. Procedure and Data Analysis
A preliminary procedure was performed to implement exploratory factor analysis (EFA). All CRPQ items were assessed for normality with skewness and kurtosis statistics. Acceptable limits of ±2 were assumed (
Field 2009).
Next, the EFA was carried out with a maximum likelihood (ML) estimation, eigenvalues > 1, and promax rotation as the factors were expected to be correlated. The authors used the Kaiser–Meyer–Olkin measure of sampling adequacy with a cut-off of 0.80 considered meritorious (
Hutcheson and Sofroniou 1999) and Bartlett’s test of sphericity value of
p < 0.05 (
Hair et al. 2019a,
2019b). Additionally, a scree plot was examined as a good graphical measure for determining the number of underlying factors. The total variance was assumed at the level of 56.6%, following Peterson’s results based on the meta-analysis of behavioral data (
Peterson 2000). The authors considered at least 5% (
Hair et al. 2019a,
2019b) for a second factor and items with loadings above 0.63 (
Tabachnick and Fidell 2013) to ensure a more reliable and robust questionnaire (
Szcześniak et al. 2022). The reliability of expected factors was measured, assuming a value of α > 0.70 as an acceptable indicator of the internal consistency of the CRPQ.
As in the final selection of the proper items, it is preferable to take into account various indicators, the next computed statistic was the corrected item–total correlation, which shows consistency between an item and other items in a factor (
Zijlmans et al. 2019). The range of item–remainder correlations (
Bandalos 2018) was considered acceptable between 0.30 and 0.70.
The current research project was approved by the Research Ethics Committee of the John Paul II Catholic University of Lublin (Institute of Theological Sciences), number KEBN_31/2022. All statistical analyses were computed with the use of IBM SPSS statistics package version 20 and IBM SPSS AMOS 21.
3.4. Results
The descriptive statistics showed an approximately normal distribution as all items tested ranged between −2 and +2 for the values of skewness and kurtosis (
Table 2).
The indices of the appropriateness of the data for factor extraction confirmed the adequacy of the sample to conduct such analysis. The Kaiser–Meyer–Olkin statistics showed a value of 0.911 indicating no problem with the sample size. The Bartlett Test of Sphericity of all fifteen items displayed acceptable correlations to continue the factorial analysis (χ2 = 1817.448, df = 105,
p < 0.001). The EFA with the maximum likelihood (ML) estimation (unforced promax rotation) showed two components with eigenvalues greater than 1.0 which accounted for 58.5% of the variance. The graphical presentation (a scree plot) also suggests a two-factor solution for the CRPQ (
Figure 1).
All items showed loadings greater than 0.6, except for item CRPQ8, and a very good internal consistency (Factor 1—official religiosity α = 0.917 and Factor 2—folk practices α = 0.816). Based on this criterion, item CRPQ8 was excluded from further analysis (
Table 3). The percentage of variance explained for official religiosity was 45.797 and for folk practices was 12.752.
The corrected item–total correlations were between 0.429 and 0.746 for items of official religiosity and between 0.530 and 0.617 for items of folk practices. Therefore, to avoid the risk of redundancy, the item with the highest value of 0.746 (CRPQ14) was removed. Thus, the following items were accepted for the CFA analysis in Study 2: official religiosity—CRPQ1, CRPQ7, CRPQ22, CRPQ29, CRPQ38 and Folk practices—CRPQ2, CRPQ4, CRPQ26, CRPQ35, CRPQ39. The Cronbach alpha for official religiosity was α = 0.790, and for folk practices, it was α = 0.787.
4. Study 2
4.1. Participants
Sample 2 consisted of 181 adults between 18 and 82 (M = 28.71; SD = 11.87) and included 116 women (64%) and 65 men (36%). As Sample 2 is part of the entire pool of respondents, divided randomly for the purposes of EFA and CFA, the variables that were investigated are the same. With regards to the place of residence, 24% of respondents indicated the countryside, 15%—a city with up to 25,000 inhabitants, 12%—a city between 25,000 and 100,000 inhabitants, and 49%—a city above 100,000 inhabitants. When asked about belief in the existence of God, it was M = 7.05 (SD = 3.46). As for the relationship with the Catholic Church, it was at a level where M = 5.55 (SD = 3.78). When asked about participation in the Holy Mass, most people (38%) indicated the Sunday Mass, followed by those who go to Mass only on major liturgical feasts (22%), every day (8%), and once every few weeks (7%). Exactly 25% of the participants indicated that they never took part in Holy Mass. When asked if they belonged to any group associated with the Catholic Church, the respondents indicated: University Chaplaincy Center, Light-Life Movement, Liturgical Service of the Altar, Living Rosary, Priesthood, Religious Congregation, Evangelizing Community, Catholic School, Catholic Youth Association, Parish Pastoral Care, Diocesan Diaconia, Caritas Polska, Vincentian Marian Youth, Bible Circle, Foundation of Small Feet, Neocatechumenate, Home Church, Seminary, Religious Community, School of New Evangelization, Society of Saint Pius X, and Church Choir.
4.2. Procedure and Data Analysis
As the structural equation model has the assumption of normal distribution, before the application of CFA, the skewness and kurtosis of all ten items were tested following the guidelines of
Tabachnick and Fidell (
2013). Values ± 2 were assumed as a good approximation to normality.
In the next step, model parameters were examined in the CFA. Considering the thresholds of factor loadings reported by
Harrington (
2009, p. 23), the values “above 0.71 are excellent, 0.63 very good, 0.55 good, 0.45 fair, and 0.32 poor”. The goodness-of-fit was evaluated using several widely known and common fit indices with values of an adequate model fit: adjusted to degrees of freedom (CMIN/DF, the goodness-of-fit index (GFI ≥ 0.90)); the Tucker-Lewis index (TLI ≥ 0.90); the comparative fit index (CFI ≥ 0.90), the standardized mean square residual (SRMS ≤ 0.08), the root mean square error of approximation (RMSEA ≤ 0.08; LO ≤ 0.08; HI ≤ 0.08) (
Szcześniak et al. 2022).
Finally, the convergent validity was investigated to demonstrate whether and how the CRPQ corresponds to other measures of religiosity. To realize this goal, the Centrality of Religiosity Scale (
Huber and Huber 2012) and Multidimensional Prayer Inventory (
Laird et al. 2004) were selected. The rationale for the use of both of the above-mentioned questionnaires was that they have a multivariate structure and different dimensions. Such a solution revealed the nuances of the newly created CRPQ.
4.3. Measures
The Centrality of Religiosity Scale (CRS) is a measure of the position of religious constructs, showing how central, important, or salient the religious meanings are within personality. The questionnaire consists of 15 items grouped into 5 subscales (
Huber 2006;
Huber and Huber 2012): intellect (measuring importance and frequency of cognitive analysis of religious themes), ideology (measuring subjective belief about the real existence of the transcendent reality and the level of openness to various forms of transcendence), private practice (measuring the frequency of actual attempts to contact with transcendence and its subjective significance), religious experience (measuring how often transcendence becomes an element of one’s experience, and the person has a sense of God’s presence or workings), and public practice (measuring the frequency and subjective importance of participation in religious services). The total score is the sum of the results on the subscales and provides an overall measure of the centrality of religious meanings in personality. The response options range from 1 (not at all/never) to 5 (to a great extent/very often). High scores on each subscale signify the high efficiency of the impact of religiousness on an individual’s experiences and behavior. The scale presents satisfactory psychometric properties (
Zarzycka 2011). The reliability of the CRS in the present study was α = 0.916 for intellect, α = 0.902 for ideology, α = 0.870 for private practice, α = 0.910 for religious experience, α = 0.882 for public practice, and α = 0.957 for the overall score.
The Multidimensional Prayer Inventory (MPI,
Laird et al. 2004), in the Polish adaptation by
Zarzycka et al. (
2022) is a 15-item multidimensional self-report tool measuring types of prayer. It measures five traditional forms of prayer: adoration (e.g., I praised God), confession (e.g., I acknowledged faults and misbehavior), thanksgiving (e.g., I expressed my appreciation for my circumstances), supplication (e.g., I made various requests of God), and reception (e.g., I opened myself up to God for insight into my problems). The participants assess the extent to which they engage in particular types of prayer on a 7-point Likert scale (from 1 (never) to 7 (all of the time)). The reliability of the MPI in the current study was α = 0.870 for adoration, α = 0.938 for confession, α = 0.894 for thanksgiving, α = 0.902 for supplication, and α = 0.903 for reception.
4.4. Results
Similarly to the findings in Sample 1, the descriptive statistics displayed a close-to-normal distribution as all items scored between −2 and +2 for values of skewness and kurtosis (
Table 4).
The structure of the CRPQ was confirmed through the CFA. The loadings (
Figure 2) were between 0.56 (good) and 0.87 (excellent) for official religiosity and between 0.62 (very good) and 0.72 (excellent) for folk practices. The goodness-of-fit showed that the two-factorial model consisting of official religiosity and folk practices adequately represented the data: CMIN/DF = 2.161; GFI = 0.93; TLI = 0.92; CFI = 0.94, SRMS = 0.05; RMSEA = 0.08; LO = 0.05; and HI = 0.10. Based on the results obtained, the model was accepted in its present form. The internal reliability for official religiosity was α = 0.838, and for folk practices, it was α = 0.802.
Considering that the phenomenon of folk practices constitutes a multifaceted concept, we also tried to explore an alternative three-factor model. Although the loadings of the third factor that we called popular faith (CRPQ35 and CRPQ39) were satisfactory (between 0.66 and 0.78), the loadings of the two other factors decreased (two loadings were under 0.55). Moreover, the covariance between folk practices and popular faith exceeded 0.86, thus suggesting discriminant validity issues. The goodness-of-fit showed that the three-factorial model consisting of official religiosity, folk practices, and popular faith represented the data less adequately than the two-factorial model: CMIN/DF = 3.081; GFI = 0.87; TLI = 0.88; CFI = 0.91, SRMS = 0.07; RMSEA = 0.11; LO = 0.09; and HI = 0.13. Based on the results obtained, the two-factor model was accepted in its initial form.
Pearson’s correlation values (
Table 5) showed that both official religiosity and folk practices were positively and significantly associated with all the dimensions/overall score of the CRS and all dimensions of the MPI. Moreover, official religiosity correlated positively with folk practices at the level of
r = 0.43 ***.
Additionally, a two-tailed Fisher’s z-test was conducted (
Table 5) to examine the equality of two correlation coefficients obtained from the same sample. It allows a comparison of the values of dependencies between both CRPQ variables, i.e., the CRS and MPI variables (
Lee and Preacher 2013;
Steiger 1980).
5. Discussion
The study aimed to develop a Catholic Religious Practices Questionnaire, determine its internal structure and psychometric properties, and analyze its correlations with established measures of religiosity. To the authors’ knowledge, the present paper is the first study to present a tool designed to measure Catholic religious practices, especially in relation to both official and folk dimensions.
The psychometric analyses in Study 1, consistent with the hypothesis formulated by the authors, suggested a two-factor structure for the CRPQ and very good internal consistency for both subscales. The first subscale, official religiosity (OR), refers to the extent to which a person shows compliance with the demands of the Catholic Church, referring to practice and observance of the official teaching. This dimension also assesses the level of consent to the presence of religious symbols and teachings in the public space. Official religiosity is not only a matter of the personal importance of attending religious practices but also obedience to the teachings of the Church. Therefore, it is necessary to understand that for a Catholic person, Mass is also an opportunity to profess faith and show that they identify with the group but also express an inner acceptance of the official teaching. That is why this factor combines both behavioral (participation) and cognitive (obedience to the Magisterium of the Church; cf.
Sullivan 2002) aspects of Catholic practice. The specificity of this connection is present in one of the fundamental rules of Catholic Church theology—lex orandi lex credendi (
De Clerck 1994). The principle (“the law of prayer [is] the law of belief”) can have at least three meanings: (1) the way in which the Church prays expresses its faith; (2) liturgy is a form of expression of faith and, therefore, this form cannot be separated from the substance of faith; (3) the way of praying determines the way of believing (
Ferdek 2012). The practice of officially defined forms of prayer is, therefore, inextricably linked to the content of religious beliefs. Four items in this scale are reverse-coded in order to avoid response style bias (
Józsa and Morgan 2017;
Suárez-Álvarez et al. 2018). High scores on the OR subscale reflect intensive cognitive and behavioral appreciation of the liturgy and official teaching of the Catholic Church. Low scores are connected to poorer attendance on Sunday Mass, lower observance of the teaching of the Church, and affirmation of the laicization processes in public sphere.
The second subscale, folk practices (FP), refers to the subjective importance of customary but informal forms of prayer known as popular piety, which are widespread among Catholics. This type of devotion is culture-dependent and can vary from country to country or even region to region. The dimension of folk practices refers to the belief in the importance of spontaneous forms of popular devotion in the development of faith and its genuineness (
Roszak and Tykarski 2020). Documents of the Catholic Church define the term “popular piety” as “diverse cultic expressions of a private or community nature which, in the context of the Christian faith, are inspired predominantly not by the Sacred Liturgy but by forms deriving from a particular nation or people or from their culture” (
DPPL 2002, p. 9). Popular piety is, therefore, understood as “the form of Christianity in which devotional practices such as praying the rosary, going on pilgrimages, and venerating the Virgin Mary and the saints, occupy a central position” (
Mong 2019, p. 2). It predominantly focuses on religious activities expressed by believers on a daily basis, which forms the core of their personal faith. High scores on the FP subscale show elevated appreciation of practice of the Catholic devotions and belief in the positive impact it has on the development of faith of the individual. Low scores mean rejection of spontaneous forms of piety and demeaning its role in the profile of prayer of a Catholic person.
The analyses in Study 2 have proven the factorial and convergent validity of both subscales. The obtained levels of internal reliability (α over 0.8) make this scale suitable for individual measurement purposes (
Taber 2018). Although we also tested the three-factor model, the empirical evidence showed that the two-factor model represented a better solution. To verify nomological validity, a series of correlation analyses were conducted between the CRPQ and other measures of religiosity. The correlation matrix shows significant positive associations between both factors of CRPQ, that is CRS and MPI, which supports the initial hypotheses of this study.
The correlation level between OR and FP is moderate. It shows that, to a certain degree, they share the common scope of religious practices of Catholics but remain separate factors. It is even more visible when convergent validity is evaluated. Although both factors of CRPQ correlate positively with all dimensions and the CRS total score and MPI subscales, for the most part, the correlation coefficients differ significantly in magnitude in favor of official religiosity. In all cases, the OR subscale exhibits higher, from moderate (religious experience) to strong (public practice) correlation with all the dimensions/the overall score of the CRS. It suggests that the perception of religiosity understood as an official obligation is more closely related to the way a person sees religiosity as a central life issue. The strongest binding (0.79) connects OR with public practice (from CRS). It appears entirely intuitive that a person that perceives participation in religious services subjectively as a priority is also someone that observes participation in Mass on Sunday and so-called ‘holy days of obligation’ (
Onuoha 2019). Less obvious is the fact that private practice reveals a strong correlation (0.71) with official religiosity compared to the moderate connection of this factor with folk religiosity. Some light might be shed on it in the perspective of the fact that private practice refers to the intensity of contact with transcendence and its subjective meaning (
Zarzycka 2011). Official religiosity also has a greater correlation coefficient with the confession subscale of the MPI compared to the folk practices scale. It might be connected to the fact that the attitude represented by higher levels of OR refers to the positive perception of the sacrament of reconciliation. The latter presupposes the confession of sins. At the same time, public piety lays less emphasis on the sacramental aspect of the confession of sins and opens alternative ways for those who cannot receive the sacramental forgiveness of sins (
Healy 2014).
The connection of OR to reception (from MPI) is stronger than that of FP. It shows that passive waiting for God’s wisdom and guidance is closer to official and liturgical forms of Catholic practices. It is congruent with the fact that the official teaching of the Church emphasizes that the grace of God is something that a person receives through the offering of Christ rather than personal effort (
Murray 2020). At the same time, folk piety is more of a bottom-up form of religiosity and through this, presupposes the activity on the part of the believer activates the meriting of God’s attention, reducing the attitude of passive waiting (
Francis 2014;
Ryan 2012). In addition, the intellect subscale of CRS (that refers to the cognitive aspects of religiosity) reveals a stronger connection to OR than to FR. It shows that people preferring official, liturgical forms of Catholic practices are more eager to reflect on the theological content of their faith and are more prone to pursue religious information (
Huber and Huber 2012).
No significant differences were found in the correlation of both CRPQ subscales with thanksgiving and supplication. It may suggest that both these aspects of prayer—giving thanks and making requests to God—are equally present in the official and folk forms of Catholic religious practices.
As the scale developed in the presented study aimed to measure specifically Catholic religious practices, it adheres to the indications formulated by researchers to take into account the distinctive characteristics of a particular religious tradition.
Hill (
2013) puts it explicitly: “Researchers are encouraged to develop measures indigenous to the population or culture of interest” (p. 53). Although some items in the CRPQ, for example, fit well with other Eastern conceptions of religiosity (e.g., “The religiosity of the people is testimony to the faith of people with simple hearts”), other items reflect the religious particularity of Catholicism (e.g., “I try to attend Mass every Sunday”). By creating a scale that taps accurately into the complexity of Catholic religious practices, the authors provide an opportunity for psychologists to combine research with other social and humanities scientists (e.g., cross-cultural psychologists, sociologists, theologians) in order to reveal associations between religious behavior and a broad spectrum of psychological, social, and cultural factors. In this sense, our questionnaire follows in the line of previous tools that measured Christian religious practices, rituals, and behaviors from a Catholic perspective (
Büssing et al. 2014,
2017). It is also in line with a recent research trend to develop scales that examine religious practices and behavior within a specific religion, e.g., Islam (
Aziz et al. 2021) or Hinduism (
Jayakumar and Verma 2021). In addition, data related to the role of popular piety and social change (such as the preservation of national identity during the partitions of Poland and World War II), the overthrow of communism, and current secularization trends provide a case for studying the role of folk practices among Polish Catholics (
Roszak and Tykarski 2020).