Psychometric and Gender Invariance Analysis of the Flourishing Scale in the Malaysian Context

The Flourishing Scale is a new scale designed to measure psychological well–being. FS has been translated into more than 10 languages (e.g., Portuguese, Russian, Turki, Spanish, Egyptian, French, Chinses, Japanese, Malay, Urdu, and Persian). The psychometric analysis of the Flourishing Scale has not been explored in the Malaysian context. The aims of the current study were to examine the factor structure, reliability, concurrent validity (correlate with Satisfaction with Life Scale [SWLS] and Life Project Reflexivity Scale [LPRS]) of the Flourishing Scale (FS). The EFA was conducted to explore the factor structure of FS in the Malaysian context. Then, CFA was run to test the model fit of the FS in the Malaysian context. Test of measurement invariance was also conducted to ascertain the generalizability of the factor structure of FS across gender groups. A cross-sectional survey with 663 university students (435 female students and 228 male students) from a public university in Sabah, Malaysia was conducted. The EFA results revealed a single factor solution with a total explained variance of 68.31%. The CFA result also revealed a one-factor structure with all eight items loaded in one factor. The multi-group analysis of this model demonstrated invariance by gender. FS also demonstrated high reliability and good concurrent validity. The FS was positively and significantly correlated with Satisfaction with Life Scale scores, Life Project Reflexivity Scale score and its subscale. The results supported FS appears to be a valid measure of a flourishing state, and its utilities in the Malaysian context is proven, including gender comparisons.


Introduction
There are various ways to gather primary data either qualitative sources (i.e., field observation, interview, and informal discussion) or quantitative data sources (survey questionnaire and interview questions; Soleyew, 2019). Although each of the methods applies to different techniques, the aim is to gather research data. Most researchers prefer to collect data by using a standardized survey questionnaire which has been used and tested both to determine the reliability and validity values. The standardized questionnaire that is designed by the primary research team, will be further translated by other researchers into different languages and adapted based on the target population's culture or context. For instance, Diener et al. (2009) designed a Flourishing scale (FS) which has been widely used and translated into more than ten languages (i.e. Portuguese, Russian, Turki, Spanish, Egyptian, French, Chinses, Japanese, Malay, Urdu, and Persian). This shows that this instrument is an important and popular measure, and its popularity might also be influenced by its attractive name, "flourishing" (Schotanus-Dijkstra et al., 2016).
The FS is designed to gauge an individual's major dimension of social-psychological functioning, contribution to the well-being of others, including purpose and meaning, supportive relationships, optimism, engagement, competence, being respected, and self-acceptance. Although previous studies have shown the validity of the Flourishing Scale for several languages, The present study adds to the literature by providing evidence of the validity of the Flourishing Scale in the Malaysian Context. Besides, cultural factor (such as the relationship between text, culture, and person) has also been taken into account in the process of translating the scale into the Malay version.
Flourishing is defined as a state in which a person demonstrates higher levels of social-psychological as well as subjective well-being (Diener et al., 2009). Diener et al. (2010) theorized flourishing as the fulfilment of the needs of selfacceptance, relatedness, competence, and the possession of psychological capital. People with high in the flourishing state tend to experience higher emotional, psychological, and social well-being (Keyes, 2007).
The FS that was designed by Diener et al. (2010) has been generally used for research and clinical purposes including clinical training and well-being intervention studies. This is because of its briefness, simplicity, and inclusiveness. The 8-item FS measures individual self-perceived success in the major function of their lives including self-esteem, relationships, optimism, and purpose. It measures a single psychological well-being scale. According to Hone et al. (2014), flourishing individuals are more productive at the workplace, learn more efficiently, enjoy better social relationships, tend to help their communities, face fewer restrictions on daily activities, increase life expectancy, and enhance emotional health.
In developing the FS scale, Diener et al. (2010) involved 689 university students from various universities in Singapore and the United States as participants. The study revealed that the FS scale is compatible with other wellbeing scales (e.g., the Satisfaction with Life Scale). The onedimensional structure of FS was verified with CFA (Confirmatory Factor Analysis) and the Rasch model. Based on the literature review (e.g., Da Fonseca et al., 2015;Duan & Xie, 2016;Munoz & Nieto, 2019) revealed that CFA of the Flourishing scale is fit to the model used in their studies and there is still a need to analyze the CFA of this scale based on cultural or context of one population. For instance, the internal consistency of this scale was tested and analyzed by researchers from various cultural backgrounds, Spanish (Munoz & Nieto, 2019). Brazil (Da Fonseca et al., 2015), Chinese (Duan & Xie, 2016;Tong & Wang, 2017), Netherlands (Schotanus-Dijkstra et al., 2016), and New Zealand (Hone et al., 2014).
The FS in the Spanish version was validated by Munoz and Nieto (2019), using the Rasch analysis. The Rasch analysis can investigate many scopes such as the scale's dimensionality, each item adjustment, items internal consistency, construct validity, the differential item functioning (DIF) and the response categories functioning. Furthermore, to determine the external construct validity, correlated measures were used, which showed an adequate internal consistency and a one-dimensional scale structure. The FS scale also showed a significant relationship with another scale such as satisfaction with social support and coping scales and wellbeing scales, and they all showed good convergent validity. The FS scale was also studied by Da Fonseca et al. (2015) in the Brazilian context and they found that the FS scale showed satisfactory internal consistency of the scale with a one-factor solution. Da Fonseca's (2015) study also reported convergent validity which showed a positive relationship with the construct of positivity. The validity of the FS was also shown in the Chinese community samples by past studies (i.e., Duan & Xie, 2016;Tong & Wang, 2017).
Apart from cultures, the FS has also been psychometric analyzed in diverse populations such as student samples, a community sample, a full-time employee sample, and a nationally representative population sample. By using EFA and CFA analysis, past studies also reported a single factor structure of the scale. In addition, it also showed acceptable to excellent reliability values, ranging from .78 to .95 (Schotanus-Dijkstra et al., 2016). The first version of the FS was labeled as the Psychological Flourishing scale which consisted of 12 items (Diener & Biswas-Diener, 2008). The FS scale has been amended and the latest version of the FS consisted of eight items was known as the Psychological Wellbeing scale. The FS is a valid and reliable tool that can be used with adults with suboptimal well-being to measure social-psychological functioning. Nevertheless, its use in clinical practice and intervention studies might be arguable (Schotanus-Dijkstra et al., 2016).
The FS also showed good convergent evidence. Past studies revealed that FS correlated moderately to strong positive correlations with other psychological well-being scales (i.e., Basic Needs Satisfaction Scale and Ryff's Psychological Well-Being Scale). The FS also showed moderate to strong negative correlations with other anxiety, stress, and depression scales (Diener et al., 2009;Sumi, 2014). The convergent evidence of the FS has been strongly supported by the subjective well-being scales such as life-satisfaction scale, happiness, and positive emotions scales (Schotanus-Dijkstra et al., 2016). Munoz and Nieto's (2019) study also found a significant correlation between the FS scale and other scales (i.e., the well-being, social support, and coping scales).
Based on the preceding argument, the FS scale showed a good psychological well-being construct, and it has been widely used across different cultures and populations. Seligman (2011) considered the FS scale as a good measurement of wellbeing. In this study, the CFA of the FS scale was examined based on the Malaysian context. Besides, the concurrent validity of the FS scale with other existing well-being scales (i.e., Satisfaction with Life Scale [SWLS] and Life Project Reflexivity [LPRS] were also examined. The measurement invariance was also conducted to discover the generalizability of the factor structure of FS across gender groups.

Respondents
The current study consisted of 663 undergraduate students (228 male and 435 female) from one of the universities in Sabah, Malaysia (with ages ranging from 18 to 28 years), of all 32% was the students in their second year of study and 68% was third-year students. The sample of this study was selected using a convenience sampling method. The respondents involved in this study were the students who attended the researchers' lecture.
The sample size of this study was determined based on the recommendation of sufficient statistical power and the expense of data collection. Tabachnick et al. (2001) suggested a minimum sample size of 200 to conduct a multivariate statistical technique, and to run a confirmatory factor analysis, the ratio of cases to free parameters should be 10:1 (Kline, 2005). The Flourishing Scale contained eight constructs and there were 663 cases involved in this study, which means that the ratio of cases to free parameters was about 83:1 which indicated the size of the sample for the current study was sufficient and should not be an issue.

Instruments
The Flourishing Scale (FS). The Flourishing Scale (FS) The scale was developed to measure individual positive functioning (e.g., optimism, positive relationships, self-esteem, and purpose and meaning in life). FS comprises eight items that provide an overall well-being score. The responses for the items were based on 7-point scales (from 1 = strongly disagree to 7 = strongly agree). The respondents who gain high scores indicate that the respondent view he/her as positive terms in functioning. The examples of the items: "I am engaged and interested in my daily activities" and "I am a good person and live a good life". FS showed a high level of reliability for the Italian version (α = .88) and for the Malaysian version with α = .93 (Chua et al., 2020).
Life Project Reflexivity Scale (LPRS). Life Project Reflexivity Scale (LPRS) proposed by Di Fabio et al. (2018), it consists of 15 items to measure three dimensions: Clarity/Projectuality, Authenticity, and Acquiescencethat are generally aligned with the tenets of life and identity construction theory (Guichard, 2005). Authenticity refers to individuals' awareness of their future career-personal-life projects as a basis of authentic values. Clarity/Projectuality refers to individuals' clarity in assessing what they want to become in their next life chapters. Acquiescence refers passively accepting values imposed by society rather than basing their own career-life projects on authentic values. The items response format was a 5-point Likert scale (from 1 = strongly disagree to 5 = strongly agree). The examples of the items: "The projects for my future life are clearly defined" (Clarity/Projectuality); "The projects for my future life are full of meaning for me" (Authenticity); and "The projects for my future life are more anchored by the values of the society in which I live than my most authentic values" (Acquiescence). The Cronbach Alpha coefficient for the Malaysian version of LPRS was .87 for Clarity/Projectuality, .88 for Authenticity, .83 for Acquiescence, and .90 for the total score (Chua et al., 2020).
The Satisfaction with Life Scale (SWLS). The Satisfaction with Life Scale (SWLS) consists of five items, it is created to measure overall satisfaction with life (Diener et al., 1985). Diener et al. claimed that Satisfaction with life is a cognitive process of judgment. Based on their own criteria, the individuals assess the quality of their life. In this cognitive process, the person weighs their personal life priorities, judges them as a whole, and defines them as more or less satisfactory (Diener & Biswas-Diener, 2008). The response format for the SWLS item was a 7-point scale (from 1 = strongly disagree to 7 = strongly agree). High scores indicate high satisfaction with life. The level of reliability for the SWLS was Cronbach's alpha = .89 (Pavot & Diener, 2008) and Cronbach's alpha = .74 in the Malaysian context.
The three instruments had been back-to-back translated into Bahasa Malaysia via an English version by the researchers of this study. The instruments were presented to the respondents as a bilingual questionnaire because Bahasa Malaysia is an official language used in Malaysia and English is a second language that is widely understood.

Data Analysis
Data of the study were analyzed using the IBM SPSS AMOS 23 Program. An Exploratory Factor Analysis was conducted to determine the factor structure of the Flourishing Scale in the Malaysian context. The Confirmatory Factor Analysis (CFA) was conducted to test the single-factor model of this scale. We estimated the FS model using the maximum likelihood method.
To determine the adequacy of the model, we compared the results of the analysis to the recommended model fit indices suggested by Hu and Bentler (1998;as shown in Table 2). The chi-square value (χ 2 ) and its associated degree of freedom (df), Goodness of fit GFI), and Adjusted Goodness of Fit (AGFI) value was also reported. In addition, the reliability and validity of FS were also tested in this study.

Exploratory Factor Analysis
Prior to Exploratory factor analysis (EFA), the items of the Flourishing Scale (FS) were tested with the initial analysis, Bartlett's test of Sphericity and Kaiser-Meyer-Olkin (KMO) to test for sampling adequacy. The Kolmogorov-Smirnov test, the Shapiro-Wilk test, skewness, and kurtosis values were reported to determine the normality of the data. The outliers (the total score of FS below 20) The analysis revealed that the KMO coefficient of FS was .942 and Bartlett's test of Sphericity result was significant (χ 2 = 3,777.16, p < .001) which revealed that the sample was sufficient. Although the Kolmogorov-Smirnov test and the Shapiro-Wilk test were significant (the result may be due to the large sample size used in this study), the skewness (−0.345) and kurtosis (−0.746) values indicated that the data of this study were normally distributed. These results supported the factorability of the FS dataset (Chua et al., 2014).
In this study, Principal Component analysis was conducted using Promax Oblique with Kappa four Rotation to explore the underlying structure of FS in the Malaysian context. All eight items of The FS were analyzed for factors. The decision on the number of factors to extract was based on the criterion of eigenvalues ≥1.0. Whereas the criterion used to select items for each factor were the factor loadings of 0.50 or above within one factor (Hair et al., 1998). The results of EFA revealed that the scale only comprised of one factor with an eigenvalue of 5.47, and 68.31% of the total variance in FS. The factor loading for FS items ranged from 0.74 to 0.85.

The Reliability of the Flourishing Scale
The method of internal consistency Cronbach's alpha was used to examine the level of reliability of FS. Cronbach's alpha with a coefficient greater than .70 indicates adequate reliability (Cortina, 1993). The result indicated a high level of reliability for FS with Cronbach's alpha = .93. The result of item analysis indicated all the items of FS were in a good range (.67-.80) of the corrected item-total correlation coefficient. The Cronbach Alpha coefficient for FS in the male sample was .98 and in the female sample was .84.

The Validity of Flourishing Scale
The validity of FS was tested using concurrent validation strategy by examining the relationship between FS with the Satisfaction with Life Scale (SWLS) score and Life Project Reflexivity Scale score. We hypothesized that the FS score would be positively and significantly associated with SWLS score and LPRS score and its sub-scale.
Refer to Table 1, the FS score was positively and significantly related with SWLS score (r = .31, p < .05) and LPRS score (r = .32, p < .05). The correlation between the FS score with sub-scale of Project Reflexivity Scale range (from r = .19 to .31, p < .001. The results confirmed the concurrent validity of FS.

Confirmatory Factor Analysis (CFA)
The result of CFA for the single-factor model of the FS shown in Table 2. The analysis revealed that the FS model showed a good fit to the data, the χ 2 (df) = 180.54 (40), p < .001, CMIN/ DF ratio = 4.514, CFI = 0.963; TLI = 0.948; RMSEA = 0.073, all achieved the recommended model fit indices. Item loading for FS model indicated the items loaded sufficiently, estimates ranged from .65 (item 3) to .879 (item 8).

Gender Invariance
Test of measurement invariance was conducted to ascertain the generalizability of the FS across gender groups using multiple-group confirmatory factor analysis. The analysis was conducted to test whether the unconstrained and constrained models of FS differ based on gender (male students = 228 and female students = 435). The result showed that there were no significant differences between the unconstrained and constrained models, which indicated that the FS model was valid for male and female groups, the changes Δχ 2 (8) = 10.08, p = .259. The result also explained that gender invariance or gender was not different at the model level.

Discussion
Psychological well-being affects a person's overall quality of life, longevity, and health. Many researchers point to the fact that people with higher psychological well-being have fewer mental health issues, fewer social problems, and are involved in more prosocial behaviours (Boehm et al., 2011;Deci & Ryan, 2008;Huppert, 2009;Xu & Roberts, 2010). As psychological well-being is a greatly valued aspect of a person's life, effort should be made to develop and explore the instruments which could be used to measure psychological wellbeing. Therefore, the main objective of the present study was focused on examining the psychometric properties, that is, (1) factor structure, (2) reliability, (3) concurrent validity and discriminant validity, and (4) measurement invariances across gender lines for the Flourishing Scale in the Malaysian setting. The Flourishing Scale has eight items and the scores computed from these items provide a single psychological well-being score of an individual responding to the instrument. Concerning our first objective, which was focused on exploratory factor analysis, it was found that the scale only comprised of one factor, which is consistent with the work of Choudhry et al. (2018). While exploring the factor structure of the Urdu Flourishing Scale, they also reported that based on principal component analysis with varimax rotation, the Urdu FS had single factor loadings. Durak and Durak (2019), while exploring the Psychometric properties of the Turkish Flourishing Scale, found that a one-factor solution for the FS was also relevant in the Turkish version. Our findings were also coherent with the original study of Diener et al. (2010). In their study, Diener et al. (2010) reported that the Flourishing Scale showed one strong factor with an eigenvalue of 4.24 which also accounted for 53% of the variance in the items during principal axis factor analysis. The factor loadings ranged from 0.61 to 0.77. Therefore, it was concluded that one strong factor characterizes the Flourishing Scale. Besides Choudhry et al. (2018) and Durak and Durak (2019), our finding was also coherent with previous research finding from other countries such as Macau (Tong & Wang, 2017), Portugal (Silva & Caetano, 2013), Spain (Ramirez-Maestre et al., 2017), and Canada (Howell & Buro, 2015). Therefore, based on the findings of the present research and the work of previous researchers, it could be concluded that FS as an instrument of psychological wellbeing (is applicable, and) focuses on a single factor only.
As for determining the reliability of the scale, which was the second objective of the present study, using the method of internal consistency by determining Cronbach's alpha, it was found that the scale has adequate reliability. Our findings regarding the adequate reliability of the translated version of the scale have been supported by early researchers in different countries as well, such as Japan (Sumi, 2014). In a study by Choudhry et al. (2018), it was found that Urdu FS has Cronbach's alpha of .914. While validating the Flourishing Scale with 608 Indian adolescents, Singh et al. (2017) also found that FS has high reliability with Cronbach's alpha ranging between .80 and .95.
While responding to our third objective, which focused on determining the concurrent validity and discriminant validity, it was found that the Flourishing Scale score was significantly and positively associated with Satisfaction with Life Scale score and Life Project Reflexivity Scale score and its sub-scale. These findings are supported by the work of Singh et al. (2017), while validating the Flourishing Scale by correlating the scales with Mental Health Continuum and its factors, they found that all correlation coefficients were significant and ranged from −.25 to .87.
The fourth research objective, which was focused on exploring the measurement invariance across gender the result found that the model was valid for both genders and there was no significant invariance reported among scores obtained by males and females. This finding is also consistent with Diener et al. (2010). Even though the results indicated no significant invariance, it was found that Cronbach Alpha for male participants was relatively high compared to female participants which could be due to the difference in the dimensions of psychological well-being measured by FS. As suggested by Perez (2012), while exploring the gender differences among Filipino male and female adolescents, it was stated that there are gender differences and similarities in the dimensions of psychological well-being such as spiritual, social, and cognitive components. Thus, it could be assumed that there is no gender invariance in psychometric properties of scale but there could be difference in overall psychological well-being which could affect psychometric values to some extent.

Conclusion
In conclusion, the results of the present study demonstrate that the FS has a single-factor structure, it has high reliability and is a valid measure of psychological well-being in the Malaysian context. It has been established as a reliable and valid measure of psychological well-being in different countries like Spain, Japan, Macau, Turkey, Pakistan, India, Russia, and Iran. The way we see and understand our life experiences influences our view of our wellbeing and happiness. For instance, in the older generation, people from different socio-economics statuses, ethnicity, cultures, or religions might view their wellbeing and happiness differently. The findings suggest that the Malaysian version of FS can be used as a valid and reliable tool to gauge individuals' psychological well-being. However, as in the present study, the sample comprised of university students only, therefore we must be careful while expanding the utility of the FS Malay version to other population groups. Furthermore, it would be interesting to explore the applicability of the FS scale in the Malaysian context on other target populations such as professionals in various fields, different communities and ethnic groups, marital, and family institutions.