Validation and psychometric properties of the Arabic version of the Quality-of-Life Scale (QOLS) with a sample of general and special education teachers Authors: Bander M. Almohayya, Khalid N. Alasim Published: 2026 ABSTRACT This study was performed to validate the Arabic version of the 15-item Quality-of-Life Scale (QOLS) using a sample of 401 teachers, including 235 general education teachers and 166 special education teachers. The psychometric properties of the QOLS were evaluated through exploratory factor analysis (EFA) and confirmatory factor analysis (CFA). EFA supported a two factor solution, explaining 55.6% of the total variance, with Factor 1 reflecting personal and social well-being and Factor 2 reflecting leisure and autonomy. Internal consistency was excellent, with Cronbach’s alpha and McDonald’s omega coefficients of .91 and .92 for the total scale, respectively. Convergent validity was established via a strong positive correlation with the WHOQOL-Bref (r = .78), while discriminant validity was confirmed through a moderate negative correlation with the Health and Suffering Scale (r = –.57). CFA demonstrated acceptable model fit indices (CFI = 0.92, TLI = 0.90, RMSEA = 0.085, SRMR = 0.05), supporting the stability of the two-factor structure. Multi-CFA further indicated that measurement invariance was achieved across gender at configural, metric, and scalar levels, whereas full invariance was not supported across educational levels. The findings indicate that the Arabic QOLS is a reliable and valid instrument for assessing quality of life among physically active adult populations such as teachers. Implications for its application in educational and research contexts, as well as limitations regarding factor structure differences from the original QOLS, are discussed. Introductions Quality of life (QoL) is a multidimensional concept that can be assessed through both general and specific frameworks using various methods of measurement [37, 47, 57]. The World Health Organization defines quality of life as “… an individual’s perception of their position in life in the context of the culture and value systems in which they live and in relation to their goals, expectations, standards, and concerns” (WHO, [55], p. 1405). In addition, several factors have been identified that have key influences on QoL, such as environmental qualities, social interactions and relationships, psychological well-being, physical health, independence, and personal values. Therefore, evaluating QoL is a multifaceted task that is influenced by a wide range of interconnected elements, including individual perspectives, cultural contexts, and external factors, which underscores the need for holistic approaches to assess and enhance QoL across different populations accurately (WHO, [55]. It is crucial to use reliable and appropriate methods to assess quality of life comprehensively and accurately. For instance, Revicki et al. [48] emphasized that the most sophisticated measurement of QoL is based upon subjective experiences and personal perceptions. These concepts, which extend beyond the definition of health, are considered “… a broad range of human experiences related to one’s overall well-being based upon subjective functioning in contrast to individual expectations” (cited in [57]. A review of the literature revealed that numerous instruments have been developed to measure QoL for various research purposes, including the Medical Outcomes Study 36-Item Short-Form Health Survey (SF-36) and the World Health Organization Quality-of-Life Assessment (WHOQoL) [34, 43, 50, 57]. However, these instruments focus primarily on health and wellness, and emphasize factors such as functional restrictions, symptoms, illnesses, and treatment results [19, 26]. While health-related factors are undeniably important in evaluating QoL, some researchers have argued that instruments that address broader dimensions should also be developed, as they pointed out that health is only one of many factors that affect an individual’s quality of life overall [19, 26, 57]. An alternative tool used to assess quality of life is the Quality-of-Life Scale (QOLS) which Flanagan [23] developed to offer a more comprehensive evaluation of QoL. The QOLS consists of 15 items grouped into five key dimensions: physical and material well-being,interpersonal relationships; participation in social, community, and civic activities; personal development and achievement, and recreational activities. This scale is designed to measure QoL based upon an individual’s perception of these various aspects of life [7, 17]. Unlike many other QoL scales, the QOLS is grounded in a theoretical framework designed to capture a more holistic view of quality of life, which makes it particularly well suited for research that involves the general population. Specifically, the QOLS was originally developed for use in a North American population and has since been translated into more than 16 languages to facilitate its use in diverse cultural contexts [7, 17, 57]. Designed initially to assess quality of life in the general population, the QOLS has been used increasingly in specific contexts, including individuals with chronic illnesses [7], institutionalized elderly individuals [49], and cardiac patients [17]. This broader use reflects the complexity of measuring general quality of life across varied populations, where interpretations may differ based upon behavioral and sociodemographic factors. The QOLS was originally developed as a 15-item instrument designed to assess individuals’ satisfaction across broad, universally relevant life domains. These domains included material and physical well-being, relationships with family and friends, participation in social and community activities, personal development, fulfillment of basic environmental, and safety needs. Items were intentionally framed to reflect those aspects of everyday living that people commonly value. Thus, the instrument is suitable for general adult populations. Burckhardt et al. [9] subsequently expanded the instrument to a 16-item version by adding an item assessing independence and autonomy, a domain considered especially relevant for individuals managing long-term chronic health conditions. The modified 16-item version is used more commonly now, particularly to assess such areas as relationships/material well-being, personal/social/community involvement, and health/functioning in individuals with chronic illnesses [9, 47]. In both versions, respondents rate each item on a 7-point Likert scale. The QOLS has been used in several studies in healthy populations, chronically ill groups, and adults of all ages [7, 9, 39, 47]. However, it has not been used with children, and therefore, it is currently not recommended for younger populations [7, 47]. In addition, although the QOLS has been adapted for use with various populations, there remains a lack of extensive validation studies across diverse cultural and demographic contexts. Both versions of the QOLS have been employed based on population and purpose. The original 15-item version [22] is applied mainly in general adult and cross-cultural samples, when aiming to assess broad life satisfaction domains [7, 39, 41, 47]. In turn, the 16-item version, which involves adding an item on independence and autonomy, is commonly used with chronic illness or health-affected populations due to increased relevance for functional limitations. First, it was estimated in 1992 by Burckhardt et al. [7, 8]. Although the QOLS has been widely applied across different populations and cultural settings, several important gaps remain in the literature [7, 39, 47]. There is a lack of rigorously validated Arabic versions of the scale using contemporary psychometric approaches. In addition, previous studies have reported variability in the factor structure of the QOLS across different samples [8, 9, 42, 45, 47], suggesting that the dimensionality of the instrument may differ depending on the population and context. Furthermore, the QOLS has rarely been examined in professional populations such as teachers, whose occupational demands may uniquely shape their perceptions of quality of life [16, 36, 38]. Accordingly, the validation of the QOLS within an Arabic-speaking teacher population is warranted. Establishing its psychometric properties in this context is essential to ensure culturally appropriate measurement and to support its use in educational research and practice. Therefore, the present study aims to translate and validate the QOLS within an Arabic-speaking teacher population and to examine its psychometric properties, including factorial validity, reliability, and measurement invariance. Study framework This study focuses on four key parameters. Its primary goal is to assess the QOLS’s psychometric properties and to explore the broader aspects of QoL measurement in both general and diverse populations as reflected by recent research. The first parameter is related to the limited availability of instruments and the dearth of studies that incorporate comprehensive and general dimensions of QoL (e.g., [57]. For QoL measures to be effective, they must be both meaningful and consistent [7, 47]. As Anderson and Burckhardt [2], Zucoloto and Martinez [57], and Reeves et al. [47] emphasized, assessment instruments should possess strong psychometric properties and measure relevant aspects of QoL without overlapping with other associated health conceptions [2, 47, 57]. The second parameter concerns the inconsistencies in psychometric findings reported in several studies [7, 39, 47], which have yet to establish clear guidelines for validating the QOLS [6, 35, 47]. Factor structures have varied across studies with different samples [8, 9, 39, 42, 45] and are not consistent with recommended standards [6, 40]. Similarly, research using the 16-item QOLS typically has identified three factors [8, 9, 39, 42, 45], but these factors’ specific composition differs. While attempts have been made to evaluate the scale’s internal consistency, test–retest reliability, cross-language validity, and concurrent validity [8, 9, 39, 42, 45], the lack of a consistent factor structure calls for caution when interpreting these results. Therefore, further research is needed to better understand and strengthen the QOLS’s psychometric foundations. The third parameter pertains to the sample and the QOLS’s factorial validity and consistency across various populations, which lack strong support as reported in previous studies [8, 9, 39, 42, 45]. In relation to this parameter, this study intends to evaluate QOL scales within a key professional group that represents a distinct and under-researched population—individuals whose work involves direct human interaction, such as teachers. Unlike populations studied frequently in previous QoL research, teachers constitute a physically and mentally active group whose quality of life warrants closer attention. For example, many studies [12, 13, 16, 44] have shown that teachers’ stress, burnout, and emotional distress are related to a lower quality of life among them, which affects their daily teaching performance significantly and their students’ outcomes directly. Teachers often interact daily with students from diverse cultural, social, and academic achievement levels, which exposes them to a variety of situations and experiences that affect their quality of life by increasing their stress and anxiety [38]. Koenig et al. [36] emphasized that teaching is a highly demanding profession involving numerous responsibilities, which often leads to increased fatigue, stress, and emotional exhaustion. This makes it crucial to evaluate the QOLS for different physically active populations, such as teachers, and not simply for populations with chronic health conditions. It is noteworthy that the scale has not yet been applied to populations categorized as physically active [47]. Consequently, evidence for the scale’s factorial validity and reliability across diverse populations remains limited, as highlighted by Burckhardt et al. [8], Liedberg et al. [42], Offenbächer et al. [45], Latorre-Román et al. [39], and Reeves et al. [47]. The fourth and final parameter is to ensure that the QOLS is suitable for practical use and research purposes, and to establish this, at least three more phases are needed after exploratory factor analysis (EFA). These procedures consist of: 1) reassessing the EFA to identify items with a more stable factor structure; 2) testing the structural integrity rigorously using Confirmatory Factor Analysis (CFA), and 3) using CFA-based invariance testing to examine the scale’s equivalent measurement across different population subgroups (e.g., gender) [6, 35, 47]. This is important because interpreting group comparisons may become more difficult if measurement bias results from failing to create equal measurement qualities across various groups [6, 40, 47]. The current study employed the original 15-item QOLS for several reasons. Firstly, the 15-item version is more aligned with the conceptual purpose of the present study, which emphasized general life satisfaction rather than illness-specific functional concerns. Secondly, the original instrument has been extensively validated in general adult populations and remains appropriate when limitations associated with autonomy are not a central feature of the sample. Thirdly, use of the 15-item scale maintains content consistency with the earliest and most widely referenced psychometric work on the QOLS, ensuring greater interpretability and comparability with foundational studies. To date, the QOLS has been widely applied to healthy adult samples, individuals with chronic illnesses, and across a broad adult age range [7, 9, 39, 47]. However, there have not been any published studies that validate the measure for use with children or adolescents, and leading reviews have cautioned against its administration in younger populations, citing developmental differences in the interpretation and prioritization of quality-of-life domains as a basis for such recommendations [7, 8, 47]. Existing studies reinforce the notion that the QOLS is designed for an adult population and should undergo further development before it can be appropriately validated and extended to minors. Despite its widespread use, past research has documented variability in the QOLS factor structure, with studies describing one-factor, two-factor, and multi-factor solutions depending on sample characteristics, illness context, and cultural background. For example, several studies have identified a predominantly unidimensional structure [7, 8], whereas others-particularly those conducted in non-Western or multicultural settings-have reported two- or three-factor models reflecting distinctions among material well-being, social functioning, and personal development [39]. These differences emphasize the ongoing need for psychometric evaluation, especially within cultural contexts wherein the meaning, salience, and weighting of quality-of-life domains may vary. The current study thus contributes to ongoing validation efforts by examining the dimensionality of the 15-item QOLS within the target population. To assess whether the original 15-item QOLS met recommended model fit criteria, the study aimed to: (1) translate the QOLS into Arabic and evaluate its factorial validity, and (2) assess the measurement invariance (equal forms, loadings, and intercepts) and structural invariance (equal factor variances/covariances and equal means) of the QOLS across gender and educational level using a sample of physically active teachers. Measures The QOLS was originally developed in the United States in the mid-1970s, using the critical incident technique-eliciting significant or fulfilling life experiences from about 3,000 individuals who differed in age, ethnicity, socioeconomic class, and geographic area; special effort was made to include groups usually underrepresented, such as poor people, older persons, ethnic minorities, and rural residents [22]. Rather than striving for statistical representativeness, Flanagan tried to capture a broad range of perspectives. The original QOLS included 15 items that were grouped into five empirically derived domains: (a) material and physical well-being, (b) relationships with family and friends, (c) social, community, and civic involvement, (d) personal development and fulfillment, and (e) recreation and leisure activities. Items within each domain assess satisfaction with experiences and functioning, such as financial security, quality of social relationships, participation in community activities, achievement of personal goals, and engagement in enjoyable activities. In Flanagan's original version, items were rated on two five-point scales measuring "importance" and "needs met," though reliability statistics were not reported. Later modifications adopted a seven-point scale anchored by "delighted" and "awful" to increase sensitivity and reduce negative bias [3, 7, 10]. WHOQOL-Bref, or The World Health Organization (WHO) One item from each of the original 24 facets and two items that measure quality of life and health satisfaction overall constitute the WHOQOL-Bref, a shortened version of the WHOQOL-100 (WHOQOL Group, [56]. In addition to a global QoL metric, it assesses four QoL domains: environmental factors,social interactions; psychological well-being, and physical health. A better QoL is indicated by higher ratings on items that employ a Likert scale with different anchors (e.g., “not at all” to “an enormous amount”). For instance, this scale is used to answer the question, “How much do you like life?” According to the WHOQOL-Bref manual, domain scores are computed by multiplying the mean score by four. The tool makes use of the previous two weeks’ experiences [33]. Scale of health and suffering The Scale of Health and Suffering (HSS) is a self-report instrument used to assess perceived health-related suffering that is based upon both theory and empirical research. The scale has 20 items that are scored on a visual analogue scale (VAS) with 101 points. The items are anchored by word pairs that represent notions related to health and suffering from Eriksson’s theory [18], such as “lost grip on life–understanding about life”. While 18 items map to subdomains such as life passion, presence, relationships, freedom, and purpose, two items—"Barriers to health–Health” and “Unbearable suffering–bearable pain” — address health and suffering directly. Higher scores indicate better health and more bearable suffering, while lower levels show more suffering and health issues. Reverse scoring is not necessary for any of the items [25]. Methods The Arabic version of the QOLS was employed in the current study. The instrument was translated following a forward–backward procedure in accordance with ITC guidelines [31] and the COSMIN standards. Initially, two bilingual experts translated the original English version into Arabic, after which two additional bilingual experts back-translated it into English. Any discrepancies were carefully reviewed and resolved to ensure semantic accuracy and cultural relevance. Subsequently, three experts in special education and psychometrics evaluated the items for content validity, assessing clarity, relevance, and appropriateness, and minor linguistic refinements were incorporated into the final version. The study was approved by the Scientific Research Ethics Committee at the University of Hail (No. H-2025–899), and all participants provided informed consent prior to data collection. Data were collected over the course of one year from multiple sites in two major Saudi Arabian cities: Riyadh and Hail. In accordance with the approved IRB protocol, participants’ responses were kept confidential. A quantitative, cross-sectional research design was employed. Participants and procedure Participants To ensure a diverse and representative sample of varying ages and educational backgrounds, Arabic-speaking full-time general and special education teachers were recruited from the two cities mentioned above. An online questionnaire was developed using Google Forms and distributed via email contained a hyperlink. General education teachers and special education teachers who were teaching in general education primary schools were targeted and invited to participate in the study voluntarily. The initial items in the questionnaire verified eligibility. The participants were required to be a teacher currently teaching in a school system, physically active, and living and teaching in one of the schools within the two cities’ metropolitan areas. Participants were recruited using a convenience sampling approach from primary schools in Riyadh and Hail. Invitations to participate were distributed via email through school administrations, professional networks, and educational contacts, and included a hyperlink to an online questionnaire administered through Google Forms. Participation was voluntary, and no incentives were provided. To be eligible, participants were required to be Arabic-speaking, full-time general or special education teachers currently teaching at the primary school level, physically active, and working within the targeted cities; Riyadh, and Hail. Eligibility was verified through initial screening items embedded in the survey. Participants were excluded if they were not currently teaching, were working outside the specified locations or educational level, or did not complete the eligibility section. In addition, responses were removed during data screening if they were incomplete or demonstrated inattentive response patterns. Statistical analysis procedures The QOLS’s construct validity and model fit were evaluated using EFA and CFA. SPSS v. 25 was used for the EFA, and AMOS v. 25 was used for the CFA. The total sample was randomly split into two subsamples to allow independent validation of the factor structure. EFA was conducted on 200 cases, and CFA was performed on the remaining 201 cases. Although the QOLS was originally conceptualized as a unidimensional measure [22, 23, 23], prior validation studies have reported inconsistent factor structures across different cultural and population contexts (e.g., one-, two-, three-, and five-factor solutions). Given this lack of consensus and the absence of prior psychometric validation of the QOLS in Arabic-speaking teacher populations, an EFA was first conducted to identify the latent structure in this specific context. CFA was then used to test and validate the empirically derived model. For EFA, factor loadings of ≥ 0.40 were considered acceptable for item retention, with higher loadings indicating stronger associations with the latent construct [20]. The factor structure of the QOLS’s Arabic version was validated by using EFA to determine the underlying factor structure of the 15 items while allowing for correlations between factors. These are the two methods used in EFA to assess construct validity as referenced in ElHafeez et al., [1]. We conducted an exploratory factor analysis (EFA) to group the items according to underlying quality-of-life (QOL) dimensions, as the factorial structure of the QOLS has not previously been evaluated in teacher populations or educational contexts. Given the inconsistency of factor solutions reported in prior studies, no predetermined structure was imposed, allowing the data to inform the most appropriate dimensional model. The number of factors to be retained was determined based on multiple criteria, including eigenvalues greater than 1, inspection of the scree plot, and parallel analysis, alongside consideration of the conceptual interpretability of the factor structure. This approach is consistent with recommendations that factor structure should be empirically examined when adapting instruments across cultural and population contexts. By using the first sub sample (200 cases), Bartlett’s test of sphericity and the Kaiser–Meyer–Olkin (KMO) test were used to evaluate factorability before the EFA was performed. The KMO index measures sample adequacy, in which values nearer 1 imply greater adequacy. Bartlett’s test is used to examine whether the variables are correlated, and a significant p-value (p < 0.05) indicates that factor analysis is appropriate. Direct oblimin rotation and maximum likelihood extraction were used to determine factor loadings for the EFA [4]. Skewness values for the items in the present study ranged from approximately −0.3 to −1.3, indicating a moderate negative skew. Although some researchers recommend using polychoric correlations and Weighted Least Squares (WLS) estimation for highly skewed ordinal data, the skewness values observed here are well within the acceptable range for maximum likelihood (ML) estimation for 7-point Likert-type items (± 2; [21, 35]. The moderate degree of skewness observed here does not seriously distort the estimates of parameters, standard errors, or the model fit indices, and, therefore, ML is appropriate to use. In addition, 7-point responding allows enough gradation for the approximation of continuous measurement, again supporting the appropriateness of ML. The eigenvalue > 1 criterion, parallel analysis, and a scree plot graph were used to determine how many factors should be retained [14, 29]. Significant factors are indicated by eigenvalues greater than 1, Generally, items with factor loadings of 0.40 or higher are kept, but those with loadings of less than 0.40 are deemed inadequate [15, 20]. Further, Cronbach’s Alpha and McDonald’s Omega were computed as they are used commonly to estimate reliability. For internal consistency, Cronbach’s alpha (α) and McDonald’s omega (ω) values of ≥ 0.70 were considered acceptable, ≥ 0.80 good, and ≥ 0.90 excellent [27]. The model fit that the EFA produced was then assessed using CFA on the other subsample of 201 cases. Chi-square divided by degrees of freedom (χ2/df), standardized root means square residual (SRMR), Tucker-Lewis index (TLI), goodness of fit index (GFI), root mean square error of approximation (RMSEA), and the adjusted goodness of fit index were among the metrics used to assess the goodness of fit [28, 30]. The model fit was evaluated using commonly recommended thresholds: χ2/df values ≤ 3 indicating acceptable fit, CFI, TLI, IFI, and RNI values ≥ 0.90 indicating acceptable fit (≥ 0.95 excellent), RMSEA values ≤ 0.08 indicating acceptable fit (≤ 0.06 good), and SRMR values ≤ 0.08 indicating good fit [30, 35]. NFI, AIC, and ECVI provide complementary information on model fit, parsimony, and potential cross-sample replicability in CFA. Standardized estimates, residual moments, and modification indices were considered in the analysis to assess model fit. Convergent validity was further supported when average variance extracted (AVE) values were ≥ 0.50, and standardized factor loadings were ≥ 0.50. [24]. Lastly, the degree of agreement between the components was calculated and their correlation was ascertained using Pearson’s correlation coefficient analysis. Four types of measurement invariance were tested using a multi-group CFA: residual (or strict); metric (or weak factorial); scalar (or strong factorial), and configural invariance [11, 46, 52, 54]. Since a new model that is more constrained than the previous model is created for each form of measurement invariance, these types were examined one after the other. Measurement invariance was evaluated using multiple fit indices. Configural invariance was supported based on the following criteria: SRMR < 0.08, RMSEA < 0.08, and CFI > 0.90 [11, 30, 52]. For group sample sizes of 300 or more, measurement invariance was further assessed using the fit index change criteria recommended by Chen [11]. Changes in CFI (ΔCFI), RMSEA (ΔRMSEA), and SRMR (ΔSRMR) were examined to evaluate successive levels of invariance. Metric invariance was supported when ΔCFI ≤ 0.010, ΔRMSEA ≤ 0.015, and ΔSRMR ≤ 0.030. Scalar and residual invariance were supported when ΔCFI ≤ 0.010, ΔRMSEA ≤ 0.015, and ΔSRMR ≤ 0.010 [11, 52]. Results The participants’ demographic characteristics are presented in Table 1. The entire sample included two subsamples; general education teachers and special education teachers. Overall, 235 (58.6%) indicated that they were general education teachers and 166 (41.4%) identified themselves as special education teachers, among whom 250 (62.3%) were men and 151 (37.7%) were women. There were 186 (46.4%) teachers between 22 and 37 years old, and 215 (53.6%) between 38 and 69 years old. With respect to the educational level variable, 268 (66.8%) held Bachelor’s degrees, 60 (15%) held Master’s degrees, and 73 (18.2%) held a Ph.D. The participants’ teaching experience overall ranged from 0 to 36 years (mean = 12.85, SD = 8.36). A total of 201 (50.1%) had 1 to 20 years of experience, and 200 (49.1%) had 21 to 40 years of experience. Additional demographic characteristics of the participants are also presented in the Table 1.Table 1Demographic characteristics of the participantsVariableCategoryFrequency (n)Percentage (%)Teacher TypeGeneral Education23558.6Special Education16641.4GenderMen25062.3Women15137.7Age (years)———Mean (SD)39.03 (8.54)—Range22–69—22–3718646.438–6921553.6Educational LevelBachelor’s26866.8Master’s6015.0Ph.D7318.2Years of Experience———Mean (SD)12.85 (8.36)—Range1–40—1–20 years20150.121–40 years20049.9 Table 1Demographic characteristics of the participantsVariableCategoryFrequency (n)Percentage (%)Teacher TypeGeneral Education23558.6Special Education16641.4GenderMen25062.3Women15137.7Age (years)———Mean (SD)39.03 (8.54)—Range22–69—22–3718646.438–6921553.6Educational LevelBachelor’s26866.8Master’s6015.0Ph.D7318.2Years of Experience———Mean (SD)12.85 (8.36)—Range1–40—1–20 years20150.121–40 years20049.9 Demographic characteristics of the participants Descriptive statistics As described in the statistical analysis procedures, descriptive analyses were first conducted. For the scales’ scores, the means and standard deviations of the data gathered from the full sample (N = 401), as well as the frequencies and percentages of all variables, were calculated using descriptive statistical analyses. The mean of the entire sample’s scores was 5.47 with a SD of 0.90. The mean scores of the general education teachers’ responses were nearly the same, 5.51 with a SD of 0.89. The special education teachers’ mean responses were 5.42 with a SD of 0.90. A 78.14% approximation rate to the maximum score indicates a fairly good quality of life. Assumption checking and choice of statistical procedures Although item-level distributions exhibited moderate departures from normality, these deviations remained within ranges for which parametric methods using maximum likelihood estimation are generally considered robust [21, 35]. Nevertheless, nonparametric tests were employed for analyses involving demographic variables as a conservative analytic choice, given the ordinal nature of Likert-type responses and the presence of skewed item distributions. The use of nonparametric procedures is appropriate in such contexts, as they do not rely on strict distributional assumptions and provide a cautious approach when normality assumptions are not fully met. Nonparametric group comparisons Consistent with the nonparametric procedures outlined in the Method, nonparametric tests were conducted for demographic group comparisons. Mann–Whitney U tests indicated no statistically significant differences in QOLS scores between participants based on age grouping, as reflected by comparable mean ranks for general education teachers (Mean Rank = 200.39) and special education teachers (Mean Rank = 201.53), p > 0.05. Similarly, no statistically significant differences were observed between general education teachers (Mean Rank = 207.26) and special education teachers (Mean Rank = 192.14), p > 0.05, indicating comparable distributions of QOLS scores across teaching roles. In contrast, a Mann–Whitney U test examining gender differences revealed a statistically significant difference in QOLS scores between males (Mean Rank = 192.36) and females (Mean Rank = 160.67), p < 0.05, suggesting that QOLS score distributions differed by gender. Additionally, a Kruskal–Wallis test was conducted to assess differences in QOLS scores across educational level groups. The results indicated no statistically significant differences in mean ranks across educational levels, H = 0.64, p > 0.05, suggesting that educational level was not associated with variations in QOLS scores. While these analyses confirm the consistency of QOLS scores across demographic subgroups, the focus of this study remains its psychometric evaluation of the Arabic version of QOLS, including EFA, CFA, reliability, and measurement invariance. Exploratory factor analysis Following completion of descriptive analyses, assessment of data normality, and nonparametric group comparisons, exploratory factor analysis (EFA) was conducted on the first subsample (n = 200) to examine the underlying factor structure of the QOLS in the present sample. Before carrying out the EFA, Kaiser–Meyer–Olkin (KMO) measurements of sampling adequacy were evaluated. The overall value of the KMO index, 0.91, is indicative of excellent adequacy for carrying out the EFA [32]. The individual values for the overall MSA were all above the minimum level of 0.60 and ranged between 0.88 and 0.95, suggesting that there is no need to discard any items because all had shared ample common variance. A preliminary EFA using the oblimin method of rotation, involving the first 15 items of the Quality-of-Life Scale (QOLS), in a subsample of 200, yielded two factors. A test for the overall Kendall’s Bartlett's test of the sphericity of the correlation matrix resulted in χa(105) = 1561.76, p < 0.001, which is indicative of adequacy for EFA. The eigenvalue analysis provided evidence in support of retaining two factors. The unrotated component analysis provided the first component eigenvalue of 7.21, accounting for 48.0% in the total variance. After the rotation, the first factor accounts for 35.2% (as shown in Table 2). The findings provided support for the multidimensional nature of the Quality-of-Life Scale.Table 2Component characteristicsUnrotated solutionRotated solutionEigenvalueProportion varCumulativeSumSq. LoadingsProportion varCumulativeComponent 17.2070.4800.4805.4080.3610.361Component 21.1270.0750.5562.9250.1950.556 Table 2Component characteristicsUnrotated solutionRotated solutionEigenvalueProportion varCumulativeSumSq. LoadingsProportion varCumulativeComponent 17.2070.4800.4805.4080.3610.361Component 21.1270.0750.5562.9250.1950.556 Component characteristics Items were assigned to factors based on their primary loadings; the value of approximately 0.40 or higher was considered meaningful for interpretation. Item nine cross-loaded on both factors with some conceptual overlap; however, the difference in the loading was good enough to keep the item based on its theoretical relevance. This is discussed further in the discussion section. Overall, the pattern of factor loadings was interpretable and supported the conceptual framework of the instrument; this supports the adequacy of the two-factor solution. To determine the number of factors to retain, Horn's Parallel Analysis (PA) was conducted. Observed eigenvalues from the actual data were compared with mean eigenvalues derived from randomly generated datasets of equivalent size. Following the PA criterion, factors were retained only if the observed eigenvalue exceeded the corresponding simulated eigenvalue. The analysis indicated that Components 1 and 2 should be retained as primary factors because their observed eigenvalues were significantly larger than the simulated eigenvalues: Component 1: λ = 6.68 > 0.64; Component 2: λ = 0.53 > 0.42. Component 1 accounted for 48.0% of the total variance in the unrotated solution (sum of squared loadings = 5.41, 36.1% in the rotated solution), while Component 2 accounted for 7.5% of the total variance, sum of squared loadings = 2.93, 19.5% in the rotated solution. The two factors combined explained 55.6% of the total variance in both the unrotated and rotated solutions. EFA using oblimin rotation yielded a two-factor solution with a coherent and interpretable structure. The first factor included 10 items (3, 4, 5, 6, 7, 9, 10, 11, 12, 13), whose standardized loadings ranged between 0.46 and 0.83. These items reflect perceptions of personal satisfaction, social relationships, emotional well-being, and engagement in meaningful life roles. The highest loadings are for 7 (0.83), 5 (0.79), 6 (0.78), and 4 (0.74). This indicates that this factor is strongly defined by the items that capture fulfillment, interpersonal connection, and perceived quality of daily functioning. Although item 9 also loaded on Factor 2 (0.41), its primary loading on Factor 1 (0.46) and conceptual alignment with personal and social well-being justified retaining this item within this factor. These items collectively signal a broad dimension of overall personal and social quality of life, suggesting that higher scores on this factor signal greater perceived well-being across relational and psychosocial domains. The second factor was made up of five items (1, 2, 8, 14, 15), whose loadings ranged between 0.50 and 0.89. The items reflect satisfaction with leisure activities, autonomy in using time, and opportunities for rest and enjoyment. The highest loadings are for 15 (0.89) and 14 (0.73), indicating that this factor is particularly well-defined by perceptions of sufficient leisure time and personal freedom. In summary, Factor 2 reflects a distinct dimension of quality of life related to leisure, distinguishing discretionary and restorative aspects of well-being from broader social and personal functioning. Only the primary factor loadings for items on their strongest respective factors are shown in Table 3 for clarity and interpretability. More specifically, loadings ≥ 0.50 were highlighted for those factors with stronger loadings, while loadings ≥ 0.40 were reported for those factors that contributed slightly less variance. Items with a weaker loading or cross-loading were excluded from the table, but they were considered in all the analyses, including parallel analysis, oblique rotation, and subsequent CFA. This approach follows the recommended psychometric practices because such an approach enables the provision of an easily understandable and meaningful factor structure without compromising any important analytic information [35, 51].Table 3Factor Loading and ComponentItemFactor 1Factor 2MeanSDItem 70.8335.801.15Item 50.7945.701.37Item 60.7815.751.20Item 40.7375.761.45Item 130.7025.86.99Item 30.6956.131.12Item 100.6915.851.11Item110.6465.641.18Item 120.5815.611.28Item 90.4655.521.20Item 150.8924.801.67Item 140.7345.231.43Item 20.6095.541.17Item 80.5475.071.44Item 10.5064.611.61 Table 3Factor Loading and ComponentItemFactor 1Factor 2MeanSDItem 70.8335.801.15Item 50.7945.701.37Item 60.7815.751.20Item 40.7375.761.45Item 130.7025.86.99Item 30.6956.131.12Item 100.6915.851.11Item110.6465.641.18Item 120.5815.611.28Item 90.4655.521.20Item 150.8924.801.67Item 140.7345.231.43Item 20.6095.541.17Item 80.5475.071.44Item 10.5064.611.61 Factor Loading and Component Confirmatory factor analysis To validate the factor structure identified in the EFA, CFA was conducted. CFA indicated that the overall fit of the hypothesized model to the data was acceptable. Incremental fit indices met or neared recommended thresholds: CFI = 0.92, TLI = 0.90, NNFI = 0.902, IFI = 0.92, and RNI = 0.92. The Normed Fit Index was slightly below the generally accepted conventional cutoff, at 0.87, but parsimony-adjusted fit was acceptable, at 0.69 for PNFI. The following incremental indices further supported model adequacy: RFI = 0.84. Absolute and residual-based fit indices further supported adequate model fit: RMSEA was 0.085, 90% CI [0.070, 0.100], an acceptable estimate considering the complexity of the model [5], the SRMR was 0.05, below the recommended cut-off point of 0.08. The GFI was 0.98 and the MFI was 0.74, indicating that the data were adequately represented. Hoelter’s critical N at α = 0.05 was 104.43 and at α = 0.01 was 114.86, thus greater than the sample size and indicating model stability. The ECVI was 1.53, suggesting reasonable cross-sample generalizability. Overall, these findings supported the adequacy of the factor structure proposed. Table 4 shows that the values of the targeted indices displayed an acceptable model fit.Table 4Values of indices for the CFAModel stepχ2dfχ2/df(TLI)(CFI)(RMSEA)(GFI)(NFI)(AIC)(SRMR)(ECVI)15 items attempt203.54832.45.90.92.08.98.878834.78.051.53 Table 4Values of indices for the CFAModel stepχ2dfχ2/df(TLI)(CFI)(RMSEA)(GFI)(NFI)(AIC)(SRMR)(ECVI)15 items attempt203.54832.45.90.92.08.98.878834.78.051.53 Values of indices for the CFA Measurement invariance Subsequently, we proceeded to assess the measurement invariance (MI) of the QOLS across the gender and educational level groups because of the variation compared to the participants’ age range. The findings of the measurement invariance test of the QOLS instrument for the gender and level of education variables are shown in Table 5. We assessed the measurement invariance of gender and educational level of the Arabic version of the QOLS using a configural invariance model. Measurement invariance across gender was tested by following a sequential multigroup CFA approach. First, the configural invariance model was tested to determine whether the basic factor structure is invariant across males and females. The configural model provided acceptable albeit mixed fit: while the comparative fit indices were moderate, CFI = 0.872, TLI = 0.838, NNFI = 0.838, NFI = 0.797, RMSEA was relatively high, 0.119, 90% CI [0.104, 0.133], p < 0.001, SRMR was 0.072, and GFI was 0.966, thus supporting the general adequacy of factor structure invariance across gender.Table 5Evidence of invariance between all types of measurement invariance for gender and educational level variables for the Arabic version of QOLSMeasurement Invarianceχ2 (df)CFITLIRMSEASRMRΔχ2 (df)ΔCFIΔTLIΔRMSEAΔSRMRGender Configura l408.98(170)0.872.838.119.072-----Metric419.61 (183).857.836.114.08210.63(13)-.015.002−0.0050.010 Scalar433.914 (196).856.846.110.08414.304 (13)-.001.010-.004.002Educational Level Configural575.802 (255).808.803.130.115----- Metric623.900 (281).806.801.131.11448.098 (26)-.002-.002.001-.001 Table 5Evidence of invariance between all types of measurement invariance for gender and educational level variables for the Arabic version of QOLSMeasurement Invarianceχ2 (df)CFITLIRMSEASRMRΔχ2 (df)ΔCFIΔTLIΔRMSEAΔSRMRGender Configura l408.98(170)0.872.838.119.072-----Metric419.61 (183).857.836.114.08210.63(13)-.015.002−0.0050.010 Scalar433.914 (196).856.846.110.08414.304 (13)-.001.010-.004.002Educational Level Configural575.802 (255).808.803.130.115----- Metric623.900 (281).806.801.131.11448.098 (26)-.002-.002.001-.001 Evidence of invariance between all types of measurement invariance for gender and educational level variables for the Arabic version of QOLS Next, metric invariance was examined by constraining factor loadings to equality across groups. The model fit remained acceptable with a small decline in the comparative fit indices: CFI = 0.857, TLI = 0.836, NNFI = 0.836, NFI = 0.775. The RMSEA showed a slight reduction, 0.114, 90% CI [0.099, 0.128], while SRMR evidenced a marginal increase to 0.082. Parsimony and cross-validation indices, PNFI = 0.675; ECVI = 2.968, reflected modest improvement relative to the configural model. In general, these findings suggest metric invariance across gender, indicating that the factor loadings are acceptably equivalent for males and females. Scalar invariance was subsequently evaluated by constraining both factor loadings and item intercepts to equality. The scalar model made no meaningful changes in fit from the metric model, CFI = 0.856, TLI = 0.846, RMSEA = 0.110, SRMR = 0.084, as reflected in the trivial changes in the indices of fit, ΔCFI = −0.001. Despite some minor residual misfit, the stability of the comparative and absolute indices of fit supports scalar invariance, indicating that latent mean comparisons across gender are meaningful. Configural invariance across educational level variable was also assessed. Model fit was acceptable but below conventional thresholds for optimal fit (CFI = 0.808, TLI = 0.803), with elevated RMSEA (0.130, 90% CI [0.124, 0.154]) and borderline SRMR (0.115). These results indicate that while the same factor structure could be specified across educational groups, overall fit was limited. Following this, metric invariance testing, which constrained factor loadings to be equal across groups, led to a slight decline in the model fit indices: CFI = 0.806, TLI = 0.801, RMSEA = 0.131, and SRMR = 0.114. The difference in CFI was ΔCFI = −0.002, indicating that full metric invariance across educational levels was not supported. According to these results, partial metric invariance may be established, showing that some factor loadings vary across educational groups. Reliability analysis Following the CFA and validity analyses, internal consistency reliability was assessed using McDonald’s omega (ω) and Cronbach’s alpha (α). In the full sample, Factor 1 demonstrated high reliability (ω = 0.90, α = 0.90), while Factor 2 showed acceptable reliability (ω = 0.80, α = 0.79). The total scale exhibited excellent internal consistency (ω = 0.92, α = 0.91). In the CFA subsample, reliability estimates remained strong. Factor 1 continued to show high internal consistency (ω = 0.87, α = 0.89), and Factor 2 demonstrated adequate reliability (ω = 0.78, α = 0.81). As shown in Table 6, the total scale maintained high reliability (ω = 0.89, α = 0.91). Slightly lower omega values in the CFA sample are expected, as CFA-based reliability estimates rely on model-implied factor loadings and error variances and are therefore more conservative. Overall, the consistency of reliability coefficients across the full sample and the CFA subsample supports the stability and robustness of the scale’s internal structure.Table 6Cronbach’s Alpha and McDonald’s total values of the all responses and CFA attemptFactorCoefficient ωCoefficient αFor all responses Factor 1.90.90 Factor 2.80.79 Total.92.91For CFA Factor 1.87.89 Factor 2.78.81 Total.89.91 Table 6Cronbach’s Alpha and McDonald’s total values of the all responses and CFA attemptFactorCoefficient ωCoefficient αFor all responses Factor 1.90.90 Factor 2.80.79 Total.92.91For CFA Factor 1.87.89 Factor 2.78.81 Total.89.91 Cronbach’s Alpha and McDonald’s total values of the all responses and CFA attempt Convergent and discriminant validity Following the assessment of internal consistency reliability, convergent and discriminant validity were examined. We assessed the values of Factors 1, and 2 using the AVE to determine the convergent validity. All of the AVE values met the threshold of 50. According to Fornell and Larcker [24], the AVE is acceptable when it is equal to or greater than 0.50. In addition, a Pearson’s correlation coefficient analysis was used to determine the correlations between the Arabic version of the QOLS and the WHOQOL-BREF measure to evaluate the convergent validity. The WHOQOL-BREF measure and the Arabic version of QOLS showed a strong and positive correlation (r = 0.78, p = 0.001), as shown in Table 7. The Arabic version of the QOLS was also evaluated for correlations with the HSS to evaluate discriminant validity. The negative correlation between the QOLS and the HSS (r = –0.57, p = 0.001) demonstrated the discriminant validity of the Arabic version of the QOLS.Table 7Correlations among the WHOQOL-BREF with scales included in the analyses (N = 232)MeasuresHSSQOLSWHOQOL-BREFHSS1-.57**-.56**QOLS-.57**1.78**WHOQOL-BREF-.56**.78**1The Quality-of-Life Scale (QOLS), WHOQOL-BREF Scale, The Health and Suffering Scale (HSS)***p < 0.001; **p < 0.01; *p < 0.05 Table 7Correlations among the WHOQOL-BREF with scales included in the analyses (N = 232)MeasuresHSSQOLSWHOQOL-BREFHSS1-.57**-.56**QOLS-.57**1.78**WHOQOL-BREF-.56**.78**1The Quality-of-Life Scale (QOLS), WHOQOL-BREF Scale, The Health and Suffering Scale (HSS)***p < 0.001; **p < 0.01; *p < 0.05 Correlations among the WHOQOL-BREF with scales included in the analyses (N = 232) The Quality-of-Life Scale (QOLS), WHOQOL-BREF Scale, The Health and Suffering Scale (HSS) ***p < 0.001; **p < 0.01; *p < 0.05 Discussion These findings of our study, therefore, based on our sample of teachers, demonstrate that the QOLS represents a measure with good psychometric properties for use in Arabic cultural settings. Further, the main findings of the consistency reliability tests performed in this study are an indication that the Arabic QOLS is a reliable scale for use with teachers in Saudi Arabia. The CFA was used to confirm the factor structure identified by the EFA. The goodness-of-fit indices indicated that the two-factor model provided an acceptable fit to the data. Based upon these findings, the QOLS’ reliability was deemed strong. To the best of our knowledge, this is the first study that reported EFA, CFA, and MI in the context of international validation using an Arabic population for the QOLS. The convergent validity of this Arabic scale was assessed by AVE. More importantly, the WHOQOL-BREF and HSS measures in this study shared correlations with the Arabic version of QOLS as expected, thus indicating convergent validity. The discriminant validity of QOLS is supported by negative correlation between QOLS and the HSS. For the dimensionality and factor solution parameter, a two-factor solution was found to fit the data and the factor analysis best, but did not group the same as the original instrument. The retention of a two-factor solution was theoretically and empirically justified, as it was supported by multiple statistical criteria (eigenvalues, scree plot, and parallel analysis) and yielded a conceptually coherent structure consistent with the multidimensional nature of quality of life. This finding is consistent with previous studies reporting variability in QOLS factor structures across samples and cultural contexts, suggesting that the dimensionality of quality of life may depend on population characteristics. Our objective in this study was to evaluate the construct validity, internal consistency, and structural validity of QOLS by using a sample of different groups of teachers. A two-factor model encompasses the Personal and Social Well-Being, and Leisure and Autonomy factor. The EFA results indicate a different factor number and item distribution among these domains. Although the original study did not explicitly report retained factors from a statistical factor analysis and reported 15 empirically derived life components based on survey responses, this present study of CFA results indicated a good fit of the EFA domain structure. These are partially supportive of Flanagan’s original QOLS of 1982, which consisted of 15 items combined into five domains which included: (a) material and physical well-being; (b) relationships with family and friends; (c) social, community, and civic involvement; (d) personal development and fulfillment; and finally, (e) recreation and leisure activities. It is specifically noted that the Factor 1 of this present study covers the identified domains of Flanagan with regard to relationships, personal development, and social functioning, while Factor 2 matched very well with the original recreation and leisure domain. Thus, the two-factor structure in this study represents a more parsimonious organization of the QOLS items while retaining the core conceptual dimensions identified by Flanagan’s. As noted earlier, although QOLS has been suggested to be a valid and reliable scale, [8, 39], its psychometric results have often been mixed and fall short of the stipulated standards to prove scale validity [6, 35, 47]. The study further adds to inconsistent results with respect to the factor solution found compared to the original [23]. Other previous studies, for example, [8] reported a 3-factor solution. In another previous study, [47] tested the 15 and 16-item QOLS and thus reported on a five factor solution, but their CFA did not meet the recommendations of model fit. They later found a one-factor solution when EFA and covariance modeling analyses were conducted [45] also reported a three factor solution result, while [57] tested a five-factor structure and also evaluated the five-dimensional QOLS’s original structure, which demonstrated sufficient psychometric qualities. Results from our study showed some different loadings and a different number of the factor solutions. The present findings extend prior QOLS validation research by providing evidence from an Arabic-speaking teacher sample and clarifying how its factorial structure compares with earlier studies. While the original conceptualization of the QOLS [23] proposed five life domains and a multidimensional view of quality of life, subsequent psychometric studies have reported inconsistent factor structures across populations. For instance, Burckhardt et al. [8] identified a three-factor solution in a clinical sample, whereas [45] and [57] reported alternative three- and five-factor models with varying fit. [47], however, found that neither the original nor multi-factor structures achieved adequate CFA fit and instead supported a more parsimonious one-factor solution. Against this mixed evidence, the current study identified a stable two-factor structure with acceptable model fit and strong reliability. Importantly, the two-factor solution partially converges with previous findings in supporting the multidimensional nature of quality of life, while also diverging by offering a more parsimonious structure than most earlier models. This suggests that, in the present Arabic teacher sample, quality of life may be conceptualized in terms of a distinction between psychosocial functioning and discretionary or leisure-related aspects of life. Compared to prior studies, which often reported unstable CFA results or fragmented factor solutions, the present findings offer evidence of a relatively clearer factor structure and cultural coherence, which may reflect contextual and population-specific interpretations of quality-of-life domains. Further, it is worth mentioning here that the results of this study may be different because we tested the original QOLS, which was a 15-question test that assessed five conceptual domains instead of the current structure that was extended by adding one more question related to self-care. This question is supportive for studies that investigate the quality-of-life perceptions of people with chronic illnesses, which was irrelevant to the participants of our study. Therefore, the QOLS in our study has 15 components. The present findings extend prior QOLS research by demonstrating a stable two-factor structure in an Arabic-speaking teacher sample, a structure that differs from the original unidimensional conceptualization [22] as well as from the three- and five-factor solutions reported in some Western and clinical samples (e.g., [8, 45, 57]. Unlike earlier studies that reported mixed or inadequate CFA fit, the two-factor model identified in this study demonstrated acceptable fit, strong reliability, and partial to full measurement invariance across key demographic groups. These results suggest that the organization of quality-of-life domains may be culturally sensitive and context-dependent, reinforcing the importance of re-examining factorial structure when adapting instruments across languages and populations. However, the two-factor solution may be a function of and likely reflects the conceptual distinction between core psychosocial functioning and discretionary aspects of life. Specifically, the items loading on Factor 1 focus on personal satisfaction, emotional well-being, social relationships, and engagement in meaningful roles, which are at the heart of an individual's everyday experience of fulfillment and connectedness. The respondents probably interpreted these items as reflecting essential, ongoing aspects of quality of life that involve both self-perception and social interaction and therefore combined coherently. In contrast, Factor 2 was defined by items from leisure, autonomy, and personal freedom, emphasizing discretionary activities and opportunities for resting or enjoying oneself. Hence, these items probe for experiences related more to situational or context-dependent experiences rather than integral to daily social or emotional functioning. The respondents could have interpreted these items as separate from the core social and personal imperatives and placed their focus on time, choice, and personal agency available, and that is why they grouped together in this factor. Overall, a two-factor solution reflects a meaningful conceptual division: Factor 1 relates to the relational and psychosocial well-being, whereas Factor 2 captures the restorative and autonomous components of quality of life. This distinction does align with how the respondents could have differentiated between the core life satisfaction and the quality of the leisure/discretionary experiences in making judgements about their lives. In the analysis of our study, a cross-loading problem and low coefficients in item nine were discovered during the first EFA test. This item set cross-loadings on both latent factors during exploratory factor analysis, which indicates partial conceptual overlap between the constructs it represents. Such a situation is not unusual in scale adaptation studies and often appears if instruments are translated into another language and then applied within a different cultural context [53]. Importantly enough, however, the difference between the primary and secondary loadings for Item 9 was sufficient to justify retaining the item, and the item demonstrated clear theoretical relevance to the intended construct. Thus, retaining Item 9 was justified on both statistical and content validity considerations, since it captures an important aspect of the construct within the Arabic cultural and linguistic context. Overall, the pattern of factor loadings remained interpretable and coherent, and no single item featuring cross-loadings compromised the clarity or adequacy of the two-factor solution. These findings support the structural validity of the Arabic version of the scale and also point to the importance of balancing empirical criteria with theoretical and cultural relevance in cross-cultural scale adaptation. Examination of measurement invariance across gender indicated that the underlying factor structure of the scale was largely invariant across males and females. More specifically, the configural model suggested that, across gender, respondents generally interpreted the items in a similar way-a basic structure equivalence. Constrained factor loadings to equality resulted in a model with an essentially unchanged fit, which suggests that the magnitude of the relationships between items and their underlying constructs was comparable across gender. Further constraining item intercepts did not alter the model in any meaningful way, indicating that latent mean comparisons between males and females are meaningful and can be interpreted confidently. In sharp contrast, measurement invariance across educational levels was more complex. While it was possible to specify an overall factor structure across the different educational groups, full equivalence of factor loadings was not supported, which suggests that some items may be interpreted differently depending on educational background. However, this constraining of item intercepts did not substantially worsen fit and, thus, provides tentative support for comparisons of latent means across educational groups, although these results should be interpreted with caution. In sum, the findings indicate that the scale is robustly invariant across gender, and hence reliable comparisons between males and females can be made. Measurement invariance across educational levels, however, is only partly supported, with potential influence of educational context on how some items are perceived and understood. This underlines the importance of considering demographic characteristics when interpreting scores and reinforces the need for cautious application of the scale across diverse educational groups. Implications for research and practice The findings of the present study have important implications for both research and applied practice. From a research perspective, the validated Arabic version of the QOLS, with its demonstrated two-factor structure, acceptable reliability, and evidence of measurement invariance across gender, provides a psychometrically sound instrument for assessing quality of life in Arabic-speaking populations. This enables future studies to examine quality of life across different contexts and populations, as well as to conduct cross-cultural comparisons and longitudinal research. In addition, the identified factor structure offers a useful framework for further investigation of the underlying dimensions of quality of life in educational and occupational settings. From a practical perspective, the availability of a validated QOLS can support educational institutions, policymakers, and practitioners in assessing teachers’ quality of life and identifying factors that may influence their well-being, job satisfaction, and professional performance. The instrument may be used as a screening and evaluation tool to inform the development of targeted interventions, professional support programs, and policies aimed at enhancing teacher well-being. Improving teachers’ quality of life may also have broader implications for educational quality and student outcomes. Although the present study focused on teachers, the QOLS is a generic measure of quality of life that has been widely applied across diverse populations. Therefore, the validated Arabic version may also be applicable to other adult groups, such as healthcare professionals, university students, and community samples. Future research may further examine its psychometric properties across different occupational and cultural contexts. Limitation Yet, there are some limitations in the present study. Since there are no gold standard measures of quality of life, it was not possible to establish criterion validity for the scale. Another limitation relates to the use of a survey design, where responses are based on self-report and may be subject to response bias, which can affect the accuracy of the data. In addition, the use of structured and predefined response options may limit the depth of participants’ perspectives. A further limitation concerns the sampling approach. Participants were recruited through an online questionnaire, which may reflect a non-random sampling method and introduce potential selection bias. Moreover, the sample consisted only of general and special education teachers, which limits the generalizability of the findings to other populations. The results should therefore be interpreted within the context of this specific professional group and may not extend to other cultural, occupational, or demographic contexts.