H3: Psychological distress will mediate the relationships between resilience, religiosity, and the emotional outcomes of shame and guilt.
The sample consisted of 281 participants with a mean age of 24.38 years (SD = 7.81), ranging from 18 to 54 years. In terms of gender distribution, 74.02% of the participants were female (n = 208), and 25.98% were male (n = 73). Regarding nationality, the majority of the participants were Libyan (76.51%, n = 215), while 23.49% were from the UAE (n = 66). For educational attainment, 8.90% of participants had completed high school (n = 25), 79.72% were university graduates (n = 225), and 11.03% had completed postgraduate studies (n = 31).
Participants were eligible to take part in the study if they met the following inclusion criteria: (1) aged 18 years or older (2) able to read and understand Modern Standard Arabic (MSA), which is the formal written and widely understood version of Arabic used across the Arab world in education, media, and official communication, and (4) provided informed consent to participate in the study. Individuals were excluded if they (1) reported a current or past diagnosis of a severe psychiatric disorder that could impair their ability to complete the self-report measures reliably, or (2) failed to complete all the questionnaire items.
Participants were recruited through social media platforms (e.g., Facebook, Twitter) and the official websites of various Libyan and UAE communities and universities, using convenience and snowball sampling methods. The online questionnaire was accessible from July 28 to October 20, 2024. Informed consent was obtained from all participants, with confidentiality and privacy were rigorously maintained.
In line with the principles of quantitative research, all measures were self-administered in Modern Standard Arabic (MSA), and data were collected through an automated online platform. The researcher had no direct interaction with participants during the data collection process, thereby ensuring objectivity and detachment from the measurement process. This approach ensured that the instruments functioned independently of researcher influence, preserving the validity and reliability of responses.
The research protocol received approval from the researcher's Institutional Review Board, under reference number MMST/2/1/25 M, reflecting a commitment to ethical standards, transparency, and integrity throughout the research process.
A demographics questionnaire was administered to gather basic background information from each participant, including age, gender, education, and country of origin.
The Arabic-translated PFQ-2 questionnaire was used for assessing shame and guilt proneness (Harder and Greenwald, 1999). PFQ-2 consists of 16 items on a scale from 0 (never) to 4 (continuously or almost constantly). This measurement contains 10 items assessing shame-proneness (e.g., "feeling humiliated, embarrassed; feelings of blushing") and 6 items assessing guilt-proneness (e.g., "intense guilt, remorse, regret") using 5-point Likert-type responses (0 = never, 1 = rarely, 2 = some of the time, 3 = frequently but not continuously, 4 = continuously or almost continuously. A summed score was calculated from item responses, with higher scores indicating greater levels of shame (range = 0-40) and guilt (range = 0-24). The shame and guilt demonstrated good internal consistency in the current sample (Cronbach's α = 0.80; McDonald's ω total = 0.80; Cronbach's α = 0.84; McDonald's ω total = 0.85).
The Arabic version of the Brief Resilience Scale (BRS), developed by Baattaiah et al. (2023), is based on the original version by Smith et al. (2008). The BRS is a self-rated assessment used to quantify a person's ability to bounce back from and cope with health-related stressors, aiming to measure one's ability to thrive in the face of adversity. The BRS consists of six items with total scores ranging from 6 to 30. Sample items include statements such as "I am able to recover quickly from stressful experiences" and "I remain calm under pressure." Participants rate each item on a scale from 1 (strongly disagree) to 5 (strongly agree). In this study, the internal consistency of the scale was acceptable, with a Cronbach's alpha of 0.74.
The 13-item Muslim Religiosity Scale (MRS) Arabic version, developed by Al Zaben et al. (2015), is a self-reported measure assessing two dimensions of religiosity through two subscales: religious practices and intrinsic religious beliefs. The religious practices subscale includes 10 items, such as "How often do you attend group religious services for worship and prayer at a mosque or in small groups at work or in your home (i.e., obligatory prayers)." The intrinsic religious beliefs subscale consists of 3 items, such as "My religious beliefs are what really lie behind my whole approach to life." Each item is rated on a 5-point Likert-type scale. The MRS demonstrates acceptable reliability, with Cronbach's alpha values for religious practices (α = 0.60), intrinsic religious beliefs (α = 0.62), and the overall scale (α = 0.60).
Participants in this samples also completed the Depression, Anxiety, and Stress Scale (DASS-8) in Arabic. This self-administered tool, developed by Ali et al. (2022), assesses symptoms related to depression, anxiety, and stress. The scale comprises 3 items for anxiety, 3 for depression, and 2 for stress. Sample items include "I was unable to become enthusiastic about anything" (depression), "I found it difficult to relax" (stress), and "I felt scared without any good reason" (anxiety). Participants rated the applicability of each item on a 4-point Likert-type scale ranging from 0 ("did not apply to me at all") to 3 ("applied to me most of the time"). Total scores were computed by summing all item scores and dividing the total by two. Higher scores on the scale indicate a greater severity of symptoms associated with depression, anxiety, and stress. Reliability analysis showed satisfactory internal consistency, with Cronbach's alpha values ranging from depression α = 0.60, anxiety 0.64, and stress = 0.50 and α = 0.79 for the overall scale, affirming its reliability in measuring stress, anxiety, and depression.
Descriptive statistics were used to analyze the study population's age, sex, education level, and country. Means and standard deviations were calculated for continuous variables, while frequencies were reported for categorical variables.
CFA was conducted to examine the intercorrelations among the items and to determine if the PFQ-2 shame and guilt subscales exhibited an underlying two-factor structure. Due to the ordinal nature of the item response options, CFA for ordinal data was applied (Li, 2016) using the CFA function in the lavaan package (Rosseel, 2012) within the JASP statistical software. As model assumptions were violated, model fit was evaluated using diagonally weighted least squares estimation for the Root Mean Square Error of Approximation (RMSEA), Comparative Fit Index (CFI), and Tucker-Lewis Index (TLI). An acceptable fit was indicated by RMSEA values below 0.10 and CFI and TLI values above 0.90.
To examine item-level properties within the scale, Item Response Theory (IRT) was applied using the Partial Credit Model (PCM) (Wright and Masters, 1997), suitable for polytomous items with ordered response categories. PCM, an extension of the Rasch model, estimates individual threshold parameters for each item, capturing the difficulty of transitioning between response categories.
In PCM, the delta-tau parameterization was used to interpret item thresholds. Each item has a series of threshold (tau) parameters representing the points along the latent trait continuum where respondents are equally likely to choose between adjacent categories (Andrich, 2013). Items with evenly spread thresholds demonstrate an effective range across the trait continuum, while items with compact thresholds may cluster within a narrow range of the trait. Each item has four tau parameters (τ1, τ2, τ3, τ4) representing the thresholds for moving from one category to the next, typically across a Likert-type scale (e.g., from "Not at all" to "Very much").
Model fit was evaluated through Infit and Outfit statistics (Pina et al., 2005). These statistics, represented as mean square (MNSQ) values, assess the degree to which item responses align with model expectations: Infit: Information-weighted, capturing how well an item fits when responses are near the person's trait level. Outfit: Sensitive to outliers, capturing unexpected responses that deviate from expected patterns. Following Linacre (2002), ideal Infit and Outfit values range between 0.7 and 1.3 for acceptable fit. Values below 1 indicate overfit, suggesting items may be too predictable, while values above 1.3 indicate underfit, suggesting items may be noisy or misaligned with the primary construct.
Internal consistency was assessed using Cronbach's α, McDonald's ω total, and ω hierarchical. Cronbach's α, despite limitations, such as its assumption of tau-equivalence, which may lead to underestimations in congeneric scales (McNeish, 2018; Sijtsma, 2009), was included for comparison with prior studies, given its widespread use. However, α can inflate reliability estimates when scales have many or redundant items (Streiner, 2003). In contrast, McDonald's ω total and ω hierarchical provide a more accurate estimate of reliability for multidimensional scales, as they account for variance due to a primary general factor and specific factors (Revelle and Zinbarg, 2009). All reliability coefficients were calculated using the omega function in the JAMUVI statistical package.
Concurrent validity was evaluated by correlating the PFQ-2 shame and guilt subscales with depression, anxiety, stress (DASS-8), and resilience (Resilience Scale), while discriminant validity was assessed with religiosity subscales. All analyses were conducted using JAMOVI (Version 3.6.1), JASP, and SAS (Version 9.4).
We utilized SEM to examine the mediating role of psychological distress in the relationship between resilience, extrinsic and intrinsic religiosity, and the outcomes of guilt and shame. In the study model, resilience and both religiosity subscales served as the independent variables, psychological distress was the mediating variable, and guilt and shame were the dependent variables. Key indices, including the Comparative Fit Index (CFI), Tucker-Lewis Index (TLI), and Bentler-Bonett Normed Fit Index (NFI), all fit values, significantly exceed the standard 0.95 cutoff for a good fit. Additionally, the Root Mean Square Error of Approximation (RMSEA) and Standardized Root Mean Square Residual (SRMR) values and below 0.10, indicating minimal residual error and aligning with ideal fit criteria. The Goodness of Fit Index (GFI) and McDonald Fit Index (MFI) further confirm an outstanding model fit above 0.90 is generally considered acceptable (McDonald and Ho, 2002).