Translation, cross-cultural adaptation, reliability, and validity of the Persian version of the Recovery Assessment Scale-Domains and Stages (P-RAS-DS) for people with mental illness

Abstract

Introduction

Recovery-oriented care in mental health is an increasing priority internationally, including in Iran. The Recovery Assessment Scale – Domains and Stages (RAS-DS) is widely used to assess personal recovery, but no validated Persian version exists. This study aimed to translate, culturally adapt, and evaluate the psychometric properties of the Persian version of the RAS-DS (P-RAS-DS) in Iran.

Methods

The study followed Beaton’s guidelines for translation and cross-cultural adaptation. Psychometric evaluation was conducted with 373 service users diagnosed with mental illness, recruited from psychiatric outpatient clinics in Tehran through convenience sampling. A cross-sectional design was used. Reliability was assessed using internal consistency (Cronbach’s alpha), test–retest reliability (intraclass correlation coefficient, ICC), standard error of measurement (SEM), and minimum detectable change (MDC). Content validity was assessed via the Content Validity Ratio (CVR) and Content Validity Index (CVI). Construct validity was evaluated through exploratory factor analysis (EFA), and Spearman’s rho was used to examine both convergent validity with the Warwick–Edinburgh Mental Well-being Scale (WEMWBS) and the Engagement in Meaningful Activities Survey (EMAS), and inter-factor correlations.

Results

Service users had a mean age of 42.68 years (SD = 11.47) and were 41.3% female and 58.7% male. Two items were culturally modified during adaptation. The P-RAS-DS demonstrated excellent internal consistency (α = 0.940) and test–retest reliability (ICC = 0.941). SEM and MDC values were low (3.19 and 8.81, respectively), indicating good measurement precision. CVR and I-CVI values exceeded standard thresholds, supporting strong content validity. EFA extracted a four-factor structure (Self-determination and connectedness, Hope and Purpose, Personally valued activities, and Seeking support) accounting for 64.67% of the total variance. Inter-factor correlations ranged from weak to moderate (ρ = − 0.037 to 0.495), and self-determination and connectedness showed the strongest associations with other subscales and total score. Convergent validity was supported by moderate correlation with the WEMWBS (ρ = 0.582) and weak correlation with the EMAS (ρ = 0.268).

Conclusion

The P-RAS-DS is a culturally appropriate, reliable, and valid instrument for assessing recovery in Persian-speaking individuals with mental illness. It can be used in clinical practice and research to evaluate personal recovery outcomes. Future research should include confirmatory factor analysis.

Clinical trial number

Not applicable.

Supplementary Information

The online version contains supplementary material available at 10.1186/s12888-025-07526-4.

Introduction

The concept of recovery has redefined mental health care for individuals with conditions such as schizophrenia and bipolar disorder by shifting the emphasis from symptom management to a comprehensive approach that prioritizes well-being, autonomy, and social participation [1, 2]. Recovery is a multidimensional process strongly influenced by cultural values, beliefs, and social contexts [3]. Cultural factors shape how individuals define well-being, interpret symptoms, and engage with mental health services, affecting recovery trajectories [4]. Consequently, tools and interventions used to improve recovery must be examined for their validity and relevance within different cultural settings to ensure accurate and meaningful results.

Mental health service delivery in Iran has been predominantly medically oriented and hospital-based. However, there are increasing calls for a shift towards more recovery-oriented approaches that are community-based, focused upon quality of life and social participation [5].

A central element of recovery-oriented care is the use of validated instruments that can generate scores accurately reflecting individuals’ experiences of the recovery process. Measuring recovery from the perspective of the person receiving care enables the development of personalized, collaborative approaches to mental health services, which are core elements of recovery-oriented practice [6, 7]. Self-rated recovery measures provide practitioners with data that identify and prioritize the aspects of care most relevant to each individual, thereby promoting engagement through structured goal-setting and collaborative care planning [8, 9]. The Recovery Assessment Scale -Domains and Stages (RAS-DS) is one such instrument designed to generate scores that evaluate personal recovery in individuals with mental illness.

Hancock et al. (2015) developed the RAS-DS, a 38-item self-report measure that captures the subjective dimensions of recovery and supports person-centered care through individualized treatment planning [10–12]. It has been translated into 18 languages and applied in diverse clinical and community contexts, including community mental health services, inpatient psychiatric units, rehabilitation programs, and psychological interventions [11–15]. However, despite its widespread use, there is currently no validated and culturally adapted Persian version of the RAS-DS (P-RAS-DS), leaving a significant gap in recovery-oriented assessment tools available for use in Iranian mental health services.

This study had two primary objectives. The first was translating the RAS-DS into Persian and ensuring its cultural relevance for application in Iranian mental health contexts. The second was to examine the psychometric properties of the scores produced by the P-RAS-DS, including evidence of reliability and validity, to determine its suitability for evaluating personal recovery among Iranian individuals with mental illness.

Method

This cross-sectional study was conducted between May 2023 and July 2024 and consisted of two phases. In the first phase, the Recovery Assessment Scale–Domains and Stages (RAS-DS) was translated and cross-culturally adapted into Persian following Beaton’s guidelines [16]. In the second phase, the psychometric properties of the scores produced by the Persian version were evaluated. A cross-sectional design was selected because it is widely used in psychometric validation studies to examine reliability and construct validity at a single point in time within a defined population [17]. The study protocol was approved by the Research Ethics Committee of the University of Social Welfare and Rehabilitation Sciences on May 10, 2023 (approval number: IR.USWR.REC.1402.006). Informed consent was obtained from all participants, who were assured that their responses would remain confidential and be used solely for research purposes.

Phase 1: Translation and cross-cultural adaptation process

Cross-cultural adaptation stages

According to Beaton’s cross-cultural adaptation guidelines, the Persian version of the RAS-DS was adapted in six stages [16]. Written permission was obtained from the RAS-DS development team before the study.

Stage 1: Initial translation. Two translators independently translated the original English RAS-DS. Both were native Persian speakers fluent in English. One held a Ph.D. in occupational therapy, providing clinical expertise, and the other held a Ph.D. in English literature, ensuring linguistic precision. Each had over 20 years of professional experience. Their distinct backgrounds ensured clinical relevance and linguistic accuracy in the preliminary translations.

Stage 2: Synthesis of the first translation version. After a detailed review for accuracy, clarity, and cultural relevance, the two translations were compared and synthesized into a single version. The research team determined the most appropriate phrasing for each item to preserve the original intent. Two mental health occupational therapy experts reviewed the synthesized version for clinical accuracy and cultural appropriateness, and their feedback was incorporated.

Stage 3: Back translation. The synthesized Persian version was back-translated into English by two independent translation institutes. Translators, blinded to the original version, produced two back-translations that were compared with the original to identify discrepancies and ensure conceptual equivalence.

Stage 4: Expert Committee Review. An expert committee was convened to ensure the conceptual equivalence and cultural appropriateness of the translated RAS-DS. The committee comprised two nurses, two psychologists, two psychiatrists, two occupational therapists, and four service users from the target population. The diverse composition of the committee ensured that the translation process benefited from a broad range of clinical expertise, cultural knowledge, and lived experience.

Stage 5: Twenty-five service users completed the pre-final version of the P-RAS-DS and subsequently participated in cognitive interviews. Two complementary techniques were employed. In the think-aloud approach, participants expressed their interpretation of each item and explained the reasoning behind their responses. In the verbal probing approach, participants answered targeted follow-up questions regarding specific terms, item clarity, and the cultural appropriateness of the content. A single researcher conducted all interviews, audio-recorded, and transcribed them for analysis.

Stage 6: The Persian RAS-DS was finalized by an expert committee. The expert committee reviewed the results of pre-version testing and finalized the Persian RAS-DS (P-RAS-DS). Participants reported that the instrument’s items were clear and understandable, with no difficulties encountered. The committee concluded that the P-RAS-DS was suitable as the final version for use in Iranian mental health settings.

Participants

Service users were involved as experts in the review stage (Stage 4) to ensure that the translated RAS-DS reflected real-life recovery experiences and aligned with recovery-focused care principles [18]. Four service users were selected through purposive sampling, prioritizing individuals able to provide informed, reflective feedback. Selection criteria included having sufficient insight into their recovery process and the ability to engage with complex health-related concepts such as recovery-oriented care, well-being, and stigma. These abilities were assessed by reviewing participants’ prior involvement in mental health-related activities or their capacity to articulate recovery experiences during preliminary interviews.

The four participants, two men and two women aged between 28 and 45 years, had clinical diagnoses of schizophrenia (n = 2) or bipolar disorder (n = 2), which are the primary populations for which the RAS-DS is intended. Such purposive approaches are commonly used in recovery-oriented research to ensure that participants can provide relevant and reflective feedback [19]. Demographic information for these service users is presented in Supplementary Table 1. In addition, 25 service users were randomly drawn from the target population to participate in the pre-version testing stage (Stage 5).

Phase 2: Psychometric testing

Participants

Adults with mental illness were recruited from the outpatient psychiatric clinics and outpatient psychiatric rehabilitation units of two psychiatric hospitals in Tehran through convenience sampling between May 2023 and July 2024. Psychiatrists screened service users for eligibility. Inclusion criteria were a confirmed diagnosis of mental illness and willingness to participate. Exclusion criteria were returning an incomplete questionnaire or declining participation.

Following the commonly cited guideline of recruiting between five and ten participants per item for cross-cultural adaptation and validation studies [20], and given that the RAS-DS contains 38 items, the target sample size was between 190 and 380 participants. The selection of measurement properties for psychometric testing, including content validity, construct validity, internal consistency, test–retest reliability, and measurement error, was guided by the COSMIN (COnsensus-based Standards for the selection of health Measurement INstruments) framework [21].

At the start of the study, all participants completed a demographic questionnaire capturing age, gender, diagnosis, employment status, marital status, years of education, and history of electroconvulsive therapy (ECT). They then completed the P-RAS-DS, the Engagement in Meaningful Activities Survey (EMAS) [22], and the Warwick-Edinburgh Mental Well-being Scale (WEMWBS) [23] questionnaires in a single face-to-face session. For test-retest reliability, participants completed the P-RAS-DS again 48 h later via email, telephone, or in person.

Validity

To evaluate the content validity of the P-RAS-DS, 12 experts assessed each item based on its relevance to recovery assessment. Experts rated items as “essential,” “useful but not essential,” or “not necessary”, and the Content Validity Ratio (CVR) was calculated for each item using Lawshe’s (1975) method [24]. The CVR reflects the proportion of experts who judged an item to be essential. According to Lawshe’s table of critical values, the minimum acceptable CVR for 12 experts is 0.56, meaning items scoring below this threshold should be revised or removed. Higher CVR values indicate stronger expert agreement on the essential nature of an item [24].

In addition, experts rated each item’s relevance to the construct on a 4-point scale (1 = not relevant, 4 = highly relevant). The Item Content Validity Index (I-CVI) was calculated as the proportion of experts rating the item as either 3 (quite relevant) or 4 (highly relevant). An I-CVI above 0.79 indicated good relevance of an item to the measured construct, values between 0.70 and 0.79 suggested the need for revision, and values below 0.70 warranted item elimination [25, 26].

Scale-level content validity was assessed using two approaches. The average method (S-CVI/Ave) was calculated by averaging the I-CVI values across all items. The universal agreement method (S-CVI/UA) represented the proportion of items for which all experts rated relevance as 3 or 4. A score of ≥ 0.90 was considered excellent, while values of ≥ 0.80 were considered good. Because S-CVI/UA requires unanimous agreement, it is a more stringent criterion and typically produces lower values than S-CVI/Ave [27].

The construct validity was assessed through correlation analyses between the P-RAS-DS factor scores and the EMAS and WEMWBS scores. Pearson’s correlation was applied to normally distributed variables, and Spearman’s rho was used for non-normally distributed data. Correlations above 0.70 were considered strong, indicating a high association between the measures. Correlations between 0.40 and 0.69 were classified as moderate. Values below 0.40 were interpreted as weak, suggesting a low degree of association and limited overlap in what the measures capture [28].

The factor structure of the P-RAS-DS scores was examined using Exploratory Factor Analysis (EFA) with Principal Axis Factoring (PFA) and Varimax rotation. PFA was chosen because it models only the shared variance among items, making it more appropriate for identifying latent psychological constructs than Principal Component Analysis (PCA), which includes both shared and unique variance [29, 30]. The number of factors was determined using eigenvalues greater than one and inspection of the scree plot. Factor loadings below 0.30 were suppressed and excluded from interpretation.

Reliability

The Intraclass Correlation Coefficient (ICC) was calculated using a two-way mixed-effects model with absolute agreement to assess test-retest reliability. An ICC value above 0.75 indicated excellent reliability, meaning that the scores produced by the P-RAS-DS were highly stable over repeated administrations [31]. Internal consistency was evaluated using Cronbach’s alpha, with values between 0.70 and 0.95 indicative of satisfactory internal consistency. This score reflects that the items in each scale dimension measured the same underlying construct without redundancy [32]. The Standard Error of Measurement (SEM) and the Minimal Detectable Change (MDC) were calculated using the following formulas: and MDC=

SD is the baseline scores’ standard deviation, and ICC is the intraclass correlation coefficient.

Measures

The EMAS is a self-report questionnaire comprising 12 items that aim to capture meaningful participation in daily activities [22]. Service users are asked to rate the meaningfulness of their everyday activities using a five-point Likert scale for each of the 12 items. The responses are summed to generate a total EMAS score, ranging from 12 to 60, with higher scores indicating a greater sense of meaning in daily activities [33]. This questionnaire has demonstrated acceptable reliability and validity in Iranian populations [34].

The WEMWBS includes 14 items that evaluate various aspects of mental well-being, such as emotions, self-control, relationships, self-esteem, motivation, social connections, concentration, sense of worth, resilience, and creativity [23]. Responses are recorded on a Likert scale, with scores ranging from 1 to 5. The overall mental well-being score, ranging from 14 to 70, is calculated by summing the item scores, with higher scores reflecting better mental well-being [35]. This measure has been validated and shown to be reliable for assessing mental well-being in Iranian individuals [36, 37].

Results

A total of 405 individuals with mental illness were approached for eligibility in the study. Of these, 20 declined to participate, and 12 either returned incomplete questionnaires or dropped out of the study. Data analysis was conducted using IBM SPSS software (version 27.0). Data from 373 service users were analyzed. The average age of service users was 42.68 years, with the majority being male, comprising 58.7% of the sample. Table 1 summarizes the demographic and clinical characteristics of the service users.

Table 1.

Service users’ demographic data (N = 373)

Variables
Age, M ± SD	42.68 ± 11.47
Gender, N (%)
Female	154 (41.3%)
Male	219 (59.7%)
Marital Status, N (%)
Single	186 (49.9%)
Married	127 (34.0%)
Divorced	54 (14.5%)
Widow	6 (1.6%)
Employment Status, N (%)
Unemployed	75 (20.1%)
Part-time employment	189 (50.7%)
Full-time employment	84 (22.5%)
Student	25 (6.7%)
Diagnosis, N (%)
Anxiety Disorders	29 (7.7%)
Bipolar Mood Disorders	202 (54.1%)
Schizophrenia	84 (22.5%)
Personality Disorders	43 (11.5%)
Other	15 (4.0%)
Years of education, N (%)
< 12	146 (39.1%)
= 12	115 (30.8%)
> 12	112 (30.0%)
ECT history, N (%)
Yes	135 (36.1%)
No	238 (63.8%)

Translation and cross-cultural adaptation

All the terms in the P-RAS-DS were clear, straightforward, and easy to translate. However, two statements, “It is important to have fun” (Item 1) and “I continue to have new interests” (Item 4), had issues for direct translation. After discussions with the developer team to retain the concept of both statements, they were translated as and . The original developer approved the final version of the P-RAS-DS for use in cognitive interviews. Cognitive interview findings indicated that service users found the P-RAS-DS easy to complete and demonstrated appropriate comprehension of the items when prompted. Therefore, further modifications were not needed.

Validity analysis

Content validity

The content validity of the scores produced by the P-RAS-DS was evaluated by a panel of 12 experts who rated the relevance of each item. All items had I-CVI values greater than 0.79, indicating high agreement among experts regarding item relevance. For the overall scale, the S-CVI/Ave and S-CVI/UA were 0.97 and 0.81, respectively, representing excellent and good scale-level content validity [27, 38]. The CVR for the overall scale was 0.91. Given that the minimum acceptable CVR for 12 experts is 0.56, all items met or exceeded this threshold, indicating strong expert consensus on their essential nature. These findings demonstrate that the scores produced by the P-RAS-DS have good content validity. Table 2 presents the CVR for items and the I-CVI of relevance.

Table 2.

The values of I-CVI of relevance and CVR for items

Questions	I-CVI of relevance	CVR Total
1	1	1
2	1	1
3	0.83	1
4	0.83	0.83
5	1	1
6	1	1
7	1	1
8	1	1
9	1	0.66
10	1	1
11	0.83	0.83
12	0.83	1
13	1	1
14	1	1
15	0.83	0.66
16	1	1
17	0.91	0.66
18	1	1
19	1	1
20	0.83	1
21	1	0.83
22	1	1
23	1	1
24	0.83	0.83
25	1	1
26	1	1
27	1	1
28	1	1
29	0.83	0.66
30	1	0.66
31	0.83	0.66
32	1	1
33	0.83	0.66
34	1	1
35	1	1
36	1	1
37	1	0.66
38	1	1
CVR total		0.910

I-CVI: Item-content validity index; CVR: Content validity ratio

Construct validity

Exploratory factor analysis (Principal Axis Factoring with Varimax rotation) confirmed sampling adequacy (KMO = 0.925) and significant inter-item correlations (χ² = 17,612.085, df = 703, p < 0.05). Factor extraction based on eigenvalues greater than one and inspection of the scree plot (Fig. 1) indicated a four-factor solution. A clear elbow was visible after the fourth factor, suggesting that additional factors contributed little explanatory value. Together, the four factors explained 64.66% of the total variance in the rotated solution.

Fig. 1.

Scree plot for the P-RAS-DS

Factor 1, labeled Self-determination and connection, included items Q7–Q13 and Q25–Q38 and reflected domains such as self-confidence, hope, openness to new experiences, symptom management, and social connectedness. Factor 2, Hope and purpose, comprised items Q14–Q20 and Q24, capturing optimism, goal-setting, and a sense of direction. Factor 3, Personally valued activities, included items Q1–Q6, which focused on engaging in purposeful and valued activities. Factor 4, Seeking support, consisted of items Q21–Q23 and reflected awareness of one’s recovery process, including knowing when to seek help and understanding what contributes to improvement. Most item communalities exceeded 0.5, and primary factor loadings exceeded 0.6, indicating a robust underlying structure. Tables 3 and 4 present the EFA results and item groupings.

Table 3.

Results of exploratory factor analysis

Factors	Initial Eigenvalues			Extraction Sums of Squared Loadings			Rotation Sums of Squared Loadings
	Total	% of Variance	Cumulative %	Total	% of Variance	Cumulative %	Total	% of Variance	Cumulative %
Self-determination and connection (F1)	17.200	45.263	45.263	17.018	44.785	44.785	16.974	44.668	44.668
Hope and purpose (F2)	4.660	12.262	57.525	4.140	10.894	55.678	3.551	9.346	54.014
Personally valued activities (F3)	2.905	7.645	65.170	2.338	6.152	61.831	2.364	6.221	60.235
Seeking support (F4)	1.488	3.917	69.087	1.078	2.836	64.667	1.684	4.432	64.667

Loadings < 0.30 were suppressed in the rotated factor matrix

Table 4.

Grouping of the 38 items into four factors with rotated factor loadings (N = 373)

Domains and Questions	Mean (SD)	Communalities	Factors loadings
			Self-determination and connection (F1)	Hope and purpose (F2)	Personally valued activities (F3)	Seeking support (F4)
D1Q1	3.36 (0.63)	0.318			0.547
D1Q2	3.40 (0.59)	0.320			0.559
D1Q3	3.20 (0.62)	0.451			0.670
D1Q4	3.17 (0.61)	0.303			0.527
D1Q5	3.01 (0.60)	0.391			0.618
D1Q6	3.23 (0.61)	0.546			0.667
D2Q7	3.30 (0.58)	0.850	0.919
D2Q8	3.17 (0.72)	0.817	0.901
D2Q9	3.07 (0.74)	0.869	0.931
D2Q10	2.26 (0.54)	0.865	0.930
D2Q11	3.13 (0.63)	0.903	0.949
D2Q12	3.22 (0.63)	0.800	0.894
D2Q13	3.17 (0.63)	0.889	0.940
D2Q14	3.05 (0.54)	0.399		0.625
D2Q15	3.32 (0.55)	0.453		0.617
D2Q16	3.27 (0.55)	0.549		0.740
D2Q17	3.04 (0.52)	0.382		0.530
D2Q18	3.25 (0.64)	0.512		0.628*
D2Q19	3.43 (0.58)	0.546		0.725
D2Q20	3.12 (47)	0.328		0.564*
D2Q21	3.09 (0.55)	0.521				0.641
D2Q22	3.10 (0.64)	0.866				0.881*
D2Q23	3.03 (0.59)	0.326				0.415
D2Q24	3.16 (0.58)	0.329		0.546
D3Q25	3.29 (0.58)	0.610	0.780
D3Q26	3.15 (0.74)	0.601	0.771
D3Q27	3.05 (0.76)	0.844	0.918
D3Q28	3.25 (0.55)	0.744	0.860
D3Q29	3.10 (0.66)	0.911	0.954
D3Q30	3.12 (0.67)	0.893	0.944
D3Q31	3.11 (0.67)	0.902	0.947
D4Q32	3.21 (0.56)	0.888	0.941
D4Q33	3.05 (0.74)	0.747	0.855
D4Q34	3.11 (0.67)	0.805	0.893
D4Q35	3.03 (0.69)	0.724	0.847
D4Q36	3.23 (0.63)	0.798	0.884
D4Q37	3.20 (0.60)	0.790	0.884
D4Q38	3.27 (0.60)	0.784	0.875

*The items had a cross-loading ≥ 0.30 on a second factor. Their primary factor loading is reported above

Original test’s domains; D1: Doing things I value; D2: Looking forward; D3: Mastering my illness; D4: Connecting and belonging

Construct validity was further evaluated through convergent validity by comparing the P-RAS-DS with two related scales: the EMAS and the WEMWBS. Spearman correlation analysis was performed using a randomly selected subsample of service users (n = 104) who completed all three measures. This sample size meets the commonly accepted recommendation of at least 100 service users for correlation-based validation studies [39]. The total P-RAS-DS score showed a moderate correlation with the WEMWBS (ρ = 0.582, p < 0.001) and a weak but significant correlation with the EMAS (ρ = 0.268, p = 0.006), providing support for the convergent validity of the scale.

Among the four subscales, self-determination and connection (Factor 1) showed statistically significant correlations with both the WEMWBS (ρ = 0.564, p < 0.001) and the EMAS (ρ = 0.257, p = 0.009). Hope and purpose (Factor 2) showed weak, non-significant correlations with the WEMWBS (ρ = 0.139, p = 0.160) and the EMAS (ρ = 0.130, p = 0.189). Personally valued activities (Factor 3) also showed non-significant correlations with the WEMWBS (ρ = 0.127, p = 0.198) and EMAS (ρ = 0.023, p = 0.816). Lastly, seeking support (Factor 4) demonstrated weak, non-significant correlations with both the WEMWBS (ρ = 0.110, p = 0.268) and EMAS (ρ = 0.041, p = 0.683). Table 5 shows the results of the convergent validity analysis.

Table 5.

Correlations with WEMWBS and EMAS (N = 104)

P-RAS-DS factors		WEMWBS	EMAS
Self-determination and connection (F1)	Spearman’s rho	0.564	0.257
	p-value	< 0.001	0.009
Hope and purpose (F2)	Spearman’s rho	0.139	0.130
	p-value	0.160	0.189
Personally valued activities (F3)	Spearman’s rho	0.127	0.023
	p-value	0.198	0.816
Seeking support (F4)	Spearman’s rho	0.110	0.041
	p-value	0.268	0.683
Total	Spearman’s rho	0.582	0.268
	p-value	< 0.001	0.006

EMAS: Meaningful activities survey; WEMWBS: Warwick-Edinburgh mental wellbeing scale

In addition to external correlations, associations between each factor and the total score were examined to evaluate the scale’s internal structure. Spearman correlation analysis was used to examine the relationships among the four P-RAS-DS factor scores and their associations with the total score. Each factor score was positively and significantly correlated with the total P-RAS-DS score, with self-determination and connection (Factor 1) showing the strongest association (ρ = 0.887, p < 0.001), followed by hope and purpose (Factor 2) (ρ = 0.363, p < 0.001), personally valued activities (Factor 3) (ρ = 0.302, p < 0.001), and seeking support (Factor 4) (ρ = 0.272, p < 0.001). These findings confirm that all four factors contribute meaningfully to the overall recovery score.

Inter-factor correlations were generally low, indicating relative distinctiveness between domains. Hope and purpose (Factor 2) showed a moderate correlation with seeking support (Factor 4) (ρ = 0.495, p < 0.001) and a weak correlation with personally valued activities (Factor 3) (ρ = 0.131, p = 0.011), while all other inter-factor correlations were weak and non-significant (all ρ < 0.08, p > 0.05). These findings indicate that, although the factors are related to the overarching construct of personal recovery, they tap into partially independent dimensions. The correlation matrix for the P-RAS-DS is presented in Table 6.

Table 6.

Correlation matrix of factor scores and total P-RAS-DS score (N = 373)

Variable Pair	Spearman’s rho	p-value	95% CI
Self-determination and connection (F1)– Hope and purpose (F2)	0.031	0.556	[–0.074, 0.135]
Self-determination and connection (F1) – Personally valued activities (F3)	0.078	0.133	[–0.027, 0.181]
Self-determination and connection (F1) – Seeking support (F4)	0.032	0.538	[–0.073, 0.136]
Empowerment (F1) – Total Score	0.887	< 0.001	[0.863, 0.908]
Hope and purpose (F2)– Personally valued activities (F3)	0.131	0.011	[0.027, 0.232]
Hope and purpose (F2)– Seeking support (F4)	0.495	< 0.001	[0.412, 0.570]
Hope and purpose (F2)– Total Score	0.363	< 0.001	[0.268, 0.450]
Personally valued activities (F3)– Seeking support (F4)	–0.037	0.474	[–0.141, 0.068]
Personally valued activities (F3)– Total Score	0.302	< 0.001	[0.204, 0.394]
Seeking support (F4)– Total Score	0.272	< 0.001	[0.173, 0.366]

Reliability analysis

The internal consistency of the Persian RAS-DS was assessed using Cronbach’s alpha, while test-retest reliability was evaluated using the ICC. This study’s overall Cronbach’s alpha for the scale was 0.940, demonstrating satisfactory internal consistency. Subscale analyses further revealed that Cronbach’s alpha for the three factors was above 0.70, indicating a high consistency across the subscales. Additionally, the ICC was used to evaluate the reliability of repeated measurements. The ICC value for the total scale was 0.941, demonstrating excellent test-retest reliability. These combined results indicate that the Persian RAS-DS is internally consistent and reliable over time, making it a suitable tool for assessing recovery. Subscale-specific ICC and Cronbach’s alpha values are presented in Table 7.

Table 7.

Reliability analysis for P-RAS-DS (N = 373)

Factor	Number of items	Cronbach’s Alphas	Intraclass Correlation (ICC)a	95% Confidence Interval
				Lower Bound	Upper Bound
Self-determination and connection (F1)	21	0.988	0.988	0.986	0.990
Hope and purpose (F2)	8	0.842	0.833	0.804	0.858
Personally valued activities (F3)	6	0.773	0.759	0.715	0.797
Seeking support (F4)	3	0.766	0.766	0.722	0.804
Total	38	0.943	0.941	0.933	0.950

The SEM was 3.19, and the MDC calculated from it was 8.81. Although the responsiveness of a measurement tool should be understood as a highly contextualized attribute rather than a fixed property [40], this SEM reflects a low level of measurement error. It represents only a small proportion of the total possible score, indicating that observed scores are likely close to participants’ true scores. The MDC of 8.81 represents the smallest change that exceeds measurement error and can be interpreted as an actual change in recovery status with 95% confidence. These findings suggest that changes greater than 8.81 points are unlikely to result from measurement variability, supporting the instrument’s precision for clinical monitoring and research applications.

Discussion

To provide effective recovery-oriented mental health care, it is essential to have valid and appropriate tools for assessing the progress and outcomes both for individuals using mental health services and of psychiatric and psychosocial interventions. To this end, the present study sought to translate, cross-culturally adapt, and validate a Persian version of the RAS-DS. This scale evaluates recovery from severe mental illness from the service user’s perspective. The cross-cultural adaptation process addresses potential discrepancies arising from differences in cultural norms, values, and behaviors. A word-for-word translation can be misleading when interpreting and measuring subjective concepts like recovery [16].

Main findings

During the translation stage, two items (statements 1 and 4) presented challenges that appeared to stem from cultural and linguistic differences. “Having fun” and “continue to have new interests” lacked exact Persian equivalents, so we expanded their wording to ensure clarity for Persian-speaking respondents. Although these adaptations resolved the immediate ambiguity, the need for such modification suggests that these items may be more culturally sensitive and less universally interpretable than others.

In the factor analysis, Q4 (“It is important to have fun”) showed the lowest loading within the personally valued activities factor (Factor 3) (0.534), potentially reflecting subtle differences in how the concept of “fun” is expressed and valued in Persian-speaking contexts. Likewise, Q33 (“Even when I do not believe in myself, other people do “) initially emerged as part of a weak fifth factor with limited conceptual coherence before being retained in the final four-factor model, indicating that its content may function differently in this cultural setting. These patterns suggest that both items warrant closer examination in future research to ensure optimal construct alignment.

The results of the present study, which did not require large-scale cultural adaptation, are consistent with a previous study comparing the RAS-DS scores between Australian and Thai populations, which found that more than 70% of responses were similar across both groups [41].

The findings from this study affirm the validity and reliability of the P-RAS-DS scores for measuring recovery. Regarding content validity, expert ratings revealed high relevance for each item, with the I-CVI exceeding 0.79 for all items. The CVR was similarly strong, reinforcing the appropriateness of the tool for measuring recovery from the perspectives of Iranian individuals with mental illness. These findings suggest that the P-RAS-DS scores accurately reflect the constructs of personal recovery and are culturally relevant for the Iranian population.

The exploratory factor analysis of the P-RAS-DS revealed a four-factor structure, explaining 64.66% of the total variance. This structure aligns with previous studies, such as Hancock et al. (2015), which identified a four-factor model involving domains of hope, empowerment, connection, and personal control [10]. Similarly, other validations of the RAS-DS scores, including short-form adaptation and Thai translation, also supported a four-factor solution in different populations [41, 42].

Despite the alignment in the number of factors, the composition of these factors in the Iranian context reveals notable cultural distinctions. In the Australian validation, items such as “I have friends who have also experienced mental illness” and “I have friends without mental illness” loaded on the Connecting and Belonging domain [10], whereas in our data, these items loaded on self-determination and connection (Factor 1), with 0.84 and 0.88 factor loadings. This reallocation suggests that social connectedness is not viewed as a separate relational outcome but as a core component of personal agency and self-determination in the Iranian setting.

The Thai validation by Khemthong et al. (2024) also produced a four-factor structure but showed a ceiling effect, with 54% scoring above the most difficult item. Eleven items demonstrated different Differential Item Functioning (DIF) relative to the Australian sample. For example, self-acceptance (I like myself, − 0.97) and openness to new experiences (I continue to have new interests, − 0.88) were among the easiest to endorse, while varied friendship networks (I have friends who have also experienced mental illness, 0.74) and early warning sign awareness (I can identify the early warning signs of becoming unwell, 0.96) were among the hardest [41]. In Rasch analysis, positive logit values indicate a higher level of recovery needed for endorsement, and negative values reflect ease of endorsement [43].

The authors suggested that the differences in endorsement patterns between the Thai and Australian versions of the RAS-DS might reflect collectivist norms in Thailand, where close family bonds are valued more highly than broader social networks, as well as alignment with the Thai Buddhist concept of thum jai, which emphasizes acceptance of life circumstances rather than active problem-solving [44]. In our Iranian sample, early warning sign items were grouped within the seeking support factor (Factor 4), contributing meaningfully to the overall recovery construct. While both the Iranian and Thai validations embedded relational elements within broader domains, the Iranian data positioned them within self-determination and connection (Factor 1), suggesting that social connectedness is viewed as an integral part of personal agency. In the Thai context, these elements were linked more closely to acceptance and harmony.

One possible explanation for this divergence lies in differing religious-cultural orientations. In Thai Buddhism, the concept of thum jai promotes emotional acceptance and letting go of unchangeable circumstances, which may reduce emphasis on active symptom monitoring [45]. In Islam, particularly in the Iranian context, trust in God (tawakkul) is often paired with personal responsibility to act, creating a cultural expectation for self-monitoring and readiness to seek help as part of one’s duty to safeguard well-being [46]. This combination of faith and proactive engagement may explain why metacognitive awareness emerged as a distinct and strongly endorsed factor in the Iranian configuration. These findings highlight that although the RAS-DS’s core domains are transportable, the meaning and factor placement of specific items shift in ways that reflect local cultural values, help-seeking norms, and family roles [16].

Although Confirmatory Factor Analysis (CFA) is commonly used to test predefined factor structures, it was not conducted in this study. Given that the RAS-DS had not previously been validated in a Persian-speaking population, and considering the likelihood of cultural variation in the conceptualization of recovery, an exploratory approach was considered more appropriate. This decision is supported by established psychometric guidelines. As Fabrigar et al. (1999) emphasize, exploratory factor analysis is particularly useful for identifying latent constructs without imposing rigid, preexisting structures, making it well-suited for studies conducted in novel cultural contexts [29]. Similarly, Worthington and Whittaker (2006) recommend using EFA during the early stages of scale development, especially when instruments are being adapted for use in different cultural settings [47].

Regarding convergent validity, the total P-RAS-DS score demonstrated a positive and significant correlation with the WEMWBS and a weaker but statistically significant correlation with the EMAS. The modest strength of the association with EMAS likely reflects only the partial conceptual overlap between the two measures; while EMAS focuses narrowly on engagement in meaningful activities, the total P-RAS-DS score captures broader aspects of recovery that extend beyond functional engagement. At the subscale level, most correlations with the EMAS and WEMWBS were weak and non-significant, except for self-determination and connection (Factor 1), which showed significant positive associations with both measures. This pattern suggests that while the aggregated recovery score reflects a broad alignment with external measures, individual recovery domains capture more specific aspects of the construct that may not directly correspond to these instruments.

Although we anticipated that the personally valued activities factor (F3) would be more strongly correlated with EMAS due to their conceptual overlap, the observed association was negligible and non-significant. This weak association is consistent with differences in construct coverage; whereas the EMAS focuses on engagement in daily activities and sources of meaning, the P-RAS-DS emphasizes activities the person identifies as personally valued and meaningful rather than merely socially desirable. It encompasses a broader view of recovery that includes psychological well-being, hope, empowerment, and self-determination [3]. From a construct validity perspective, meaningful occupation is only one pathway through which recovery manifests, and its influence may be mediated by factors such as symptom self-management, social connectedness, and personal growth, which are not directly assessed by the EMAS [48].

Internal consistency of the P-RAS-DS scores was reported as Cronbach’s α = 0.94; this is close to the results of the original English version, for which Cronbach’s alpha was 0.96 [10]. These findings suggest that the scale’s items consistently measure the same underlying construct of recovery across cultures. Our results are also comparable to the adaptation of the short form of the RAS-DS, which reported a Cronbach’s α of 0.92 [42]. This suggests that the P-RAS-DS is a reliable tool for measuring recovery, with items on the scale consistently assessing related constructs.

Additionally, the P-RAS-DS scores showed excellent test–retest reliability (ICC = 0.94), reflecting excellent stability over time. This result is particularly important for ensuring that the P-RAS-DS scores can reliably measure changes in recovery for mental health treatment or rehabilitation, making it a valuable tool for clinical practice. A 48-hour interval between administrations was selected, as it was considered sufficient to minimize carryover effects while ensuring that participants’ recovery conditions were unlikely to change [49, 50]. The small value of SEM (3.19) indicates that the P-RAS-DS can evaluate changes precisely. The relatively low MDC (8.81) indicates that even small changes in recovery can be detected using the P-RAS-DS, making it a sensitive tool for tracking recovery progress in individuals with mental illness.

Implications

This study’s findings have several important implications for the mental health care system in Iran. The successful adaptation and validation of the P-RAS-DS scores provide Iranian practitioners with a culturally and linguistically appropriate means of assessing recovery. In clinical settings, the P-RAS-DS can be incorporated into routine assessments to generate scores that identify individual strengths and recovery needs, inform treatment planning, and monitor progress over time. For policymakers, these scores offer a standardized outcome measure to evaluate recovery-oriented programs and guide resource allocation toward services that demonstrate measurable impacts on recovery for people using their services. For researchers, the P-RAS-DS scores enable cross-cultural comparisons of recovery processes and support the evaluation of interventions designed to enhance personal recovery. Linking assessment results to individualized support strategies can help embed recovery principles into everyday service delivery and policy frameworks, promoting a more person-centered approach to mental health care.

Limitations and future research

This study relied on a convenience sample drawn from psychiatric hospitals and clinics in different regions of Tehran, which may limit the generalizability of the findings to the broader Iranian population. Participants were primarily urban service users with access to clinical services in the capital city, and their experiences may differ from those of individuals in rural areas or other provinces, where access to care, cultural norms, and perceptions of recovery may vary.

Additionally, this study did not examine potential differences in the P-RAS-DS scores across demographic or clinical subgroups such as age, gender, or psychiatric diagnosis, as this was beyond the scope of the current study. Future research should include more geographically and socioeconomically diverse samples, confirm the factor structure through confirmatory factor analysis (CFA) and item response theory (IRT) approaches to further evaluate the individual performance of items, and investigate measurement invariance across subgroups. Longitudinal studies are also recommended to examine the scale’s sensitivity to change over time, enhancing its value for clinical monitoring and program evaluation. Moreover, future research should assess reliability over intervals longer than two days to confirm the stability of the scores.

Conclusion

The validation of the P-RAS-DS provides a culturally relevant tool for assessing recovery in individuals with mental illness within the Iranian context. By enabling practitioners to assess recovery from the service users’ perspectives, the P-RAS-DS promotes person-centered care and facilitates collaborative goal-setting, which are essential in recovery-oriented practice. The tool’s high reliability and sensitivity make it suitable for tracking recovery progress and tailoring interventions, ultimately contributing to enhancing mental health services in Iran. Moreover, its significant positive correlation with well-being measures underscores its utility in aligning treatment goals with the subjective recovery experiences among Iranian service users. Nevertheless, these findings should be interpreted considering certain methodological constraints, including using a convenience sample, the absence of confirmatory factor analysis, and reliance on exploratory methods. Addressing these limitations in future research will strengthen the evidence base for the P-RAS-DS and support its broader application across diverse populations and settings.

Supplementary Information

Below is the link to the electronic supplementary material.

Abbreviations

RAS-DS

The Recovery Assessment Scale-Domains and Stages

P-RAS-DS

Persian version of the RAS-DS

EMAS

Meaningful Activities Survey

WEMWBS

The Warwick-Edinburgh Mental Well-being Scale

CVR

Content Validity Ratio

CVI

Content Validity Index

S-CVI/ Ave

Scale-level CVI Average

S-CVI/ UA

Universal Agreement

EFA

Exploratory Factor Analysis

ICC

Intraclass Correlation Coefficient

Author contributions

E.F.A: Conceptualization, Methodology, Writing – Original Draft, Project Administration. N.H: Conceptualization, Writing – Review & Editing, and Supervision. N.D: Conceptualization, Writing – Review & Editing, and Supervision. E.F: Conceptualization, Writing – Review & Editing, and Supervision. S.H.Z: Data Curation, Formal Analysis, Software. H.M: Conceptualization, Methodology, Writing– Original Draft, Supervision and funding acquisition.

Funding

This study was supported by a grant from the University of Social Welfare and Rehabilitation Sciences for the research expenses.

Data availability

Data are available upon reasonable request from the first author or/ corresponding author.

Declarations

Ethics approval and consent to participate

This study was approved by the Research Ethics Committee of the University of Social Welfare and Rehabilitation Science (IR.USWR.REC.1402.006). Each participant was informed of the purpose and requirements of the study. Consent was obtained as written approval on the ethical informed consent form. The procedures used in this study adhere to the tenets of the Declaration of Helsinki.

Consent for publication

Not applicable.

Competing interests

The authors declare no competing interests.