Introduction

The family unit is a fundamental building block of society. They also offer a deep sense of belonging, security, and emotional support. Family relationships significantly affect the well-being of both individuals and the society. Research consistently highlights their strong association with increased life satisfaction [1, 2], emotional well-being [3, 4] and resilience [5]. Family dynamics are essentially a microcosm of society; they provide a framework for people to learn how to negotiate relationships, carry out responsibilities, and exchange experiences [6].

Furthermore, there is evidence that numerous aspects of family well-being are directly correlated with family satisfaction [6, 7]. When examining the world of family life, many factors must be considered when calculating family satisfaction, and these factors display a wide range [8]. As a core concept for comprehending the qualities and behaviors that are crucial to fostering successful family life, researchers have increasingly embraced the idea of family life satisfaction, offering a holistic perspective. This idea refers to a person’s comprehensive subjective assessment of their own family life [8, 9].

The assessment of satisfaction with family life

Since the 1970s, researchers have concentrated their efforts on studying family life satisfaction. Consequently, they developed multiple scales and models to measure this across different cultures and age groups. During this period, researchers conducted studies to conceptualize and measure this structure [10, 11]. Over time, it has undergone repeated examination using diverse methodological approaches introduced the Family Adaptability and Cohesion Evaluation Scale II (FACES II), a tool designed to measure family satisfaction through dimensions such as harmony, flexibility, and communication [12]. However, discussions on the effectiveness of the model during the early 1990s led to a decrease in assessments based on dimensions [13]. A notable tool for measuring family satisfaction is the Kansas Family Life Satisfaction Questionnaire [14, 15]. The applicability of the questionnaire is unsustainable because of the essential requirement that families have at least four members [15]. The scale designed by Carver and Jones [16] to evaluate participants’ family satisfaction was criticized for being too long and difficult to complete. Later, Baracca et al. [17] developed the Adjective Family Satisfaction Scale to measure family satisfaction based on emotions and emotional experiences; however, difficulties were encountered in responding to the scale [18].

Recently, The Satisfaction with Family Life Scale (SWFLS), a scale based on the framework of The Satisfaction with Life Scale (SWLS) [19], was developed by Zabriskie and colleagues [8, 20]. Researchers have altered the SWLS for use in a variety of populations, including young people [21, 22], and in a variety of cultural contexts, including Turkish culture [23]. The SWLS [19] was modified to create SWFLS [8]. In this modified version, each of the original elements was updated, with the term “life” being altered to “family life.” As a result, five interrelated items were created. One of these items, for example, is “My family life is close to my ideals in most aspects.” The same methodology was applied to measure general satisfaction in various contexts, including academic [24], financial [25], job [26], love [27], and romantic satisfaction [28].

Similar to analyzing life satisfaction, evaluating family life satisfaction involves a comprehensive cognitive assessment of a person’s familial interactions, which is influenced by their subjective viewpoint [29]. A comprehensive assessment of family life satisfaction offers clarity and insight [30], similar to personal subjective well-being self-reports [31]. A thorough assessment of family life satisfaction should include the perspectives of every family member regarding their interactions and relationships [9, 32]. A more realistic measure of family satisfaction can be obtained by viewing family life from a global perspective [8]. The SWFLS, which takes a global approach to family life satisfaction, has been widely used across different countries, including Australia, Canada, the United States, the United Kingdom, New Zealand [8], Brazil [33], and in different languages, such as Spanish [34, 35] and Portuguese [33, 36].

Purpose of the Present Study

Despite the thriving field of family counseling and education research in Türkiye, available instruments for the assessment of family life satisfaction are limited. The Turkish adaptation of Family Satisfaction by Adjectives Scale (FSAS) by Taşdelen-Karçkay [18] primarily concentrates on the emotional aspect of family satisfaction. The SWFLS, a recent tool for measuring family life satisfaction [8], stands out for its alignment with the SWLS model. This makes it a powerful tool that enhances the cognitive aspect of comprehensive evaluation of family life satisfaction.

Global population diversity necessitates cross-culturally validated research scales and methods because there is a significant lack of comprehensive and culturally relevant evaluation tools. Taking these factors into account, this study aimed to adapt and validate the SWFLS in the Turkish cultural context. This adaptation involves modifying the established Satisfaction with Life Scale (SWLS) originally devised by Diener vd [19]. The psychometric qualities of the modified scale will next be thoroughly evaluated.

Study 1

Method

Participants and Procedure

Ed Diener was first contacted by e-mail to initiate the scale adjustment process. A revised version of the concept of ‘life’ on the scale has been submitted for his own approval, replacing it with ‘family life’. After obtaining permission to continue the adaptation study, ethical approval was obtained from the Democracy University, Türkiye (2021/03 − 02). The scale was then translated from English into Turkish under the expertise of two Turkish researchers, both of whom held doctorates in psychological counseling and guidance. Subsequently, these translated versions were reviewed by six experts in the field of guidance and psychological counseling. The same evaluation and combination process were repeated for the reverse-translation form. The final editing of the items led to the completion and finalization of the Turkish version of the scale. Following this, the scales were transformed and merged into an online format using Google Forms. The study’s test-retest reliability, convergent validity, and CFA were then evaluated in the three participant groups using online data collection techniques. The researchers distributed online forms to undergraduate and postgraduate students and their social networks using WhatsApp and Facebook. Snowball sampling was used to select volunteers from various cities, with final-year students given test-retest reliability from the URL to complete over two weeks. All participant groups read and signed an informed consent form before completing the sociodemographic data form and study instruments. The research data were obtained over a six-month period.

Before CFA, the multivariate normality of the dataset was assessed, and nine outliers were removed from the dataset. The participants in the CFA study were 646 individuals aged between 18 and 71 years. 61.6% (398) of the participants were female, 38.4% (248) were male, and the mean age was 31.71 (SDage = 11.79). Table 1 presents the demographic characteristics.

Table 1 Demographic characteristics of the CFA participant group
Full size table

Measures

Demographic information form

The study used a personal information form that included five demographic items: gender, date of birth, relationship status, educational status, and employment status.

The satisfaction with Family Life Scale (SWFLS)

The SWFLS was developed by adapting the widely used SWLS [19]. Its purpose is to assess overall family satisfaction by comparing family life circumstances with personal standards and expectations. The ‘life’ terms in the SWLS were rearranged by adding the term ‘family’ to the scale items to match the sentence structure. The scale uses a 7-point Likert format, ranging from 1 (strongly disagree) to 7 (strongly agree). By summing all the items within the SWFLS, a family life satisfaction score ranging from 5 to 35 can be calculated. Data for the validity and reliability studies of the scale were collected through a series of 15 related studies conducted by [8] and their colleagues over almost a decade. The validity and reliability assessments conducted by [8] included English-speaking participants from the United States, Canada, the United Kingdom, Australia, and New Zealand and demonstrated the effectiveness of the scale in measuring the unidimensional aspect of family life satisfaction, similar to the original general life satisfaction study. The test-retest reliability, applied at five-week intervals, was 0.87 in a study group of university students [8]. Information on the reliability and validity of the Turkish version of the scale can be found in the Results section of this study.

Sample size of the study

There have been various suggestions regarding the appropriate sample size to verify the reliability and validity of a scale. Based on the number of participants per item, it is advised that the validity and reliability study of the SWFLS, which consists of five items, should include a minimum of 15–100 people. For factor analysis, 300 individuals are a reasonable number, while 150–200 is also acceptable as a sample size [37, 38]; however, there has been a current effort to obtain a total number of participants close to 1000 [39,40,41,42]. We met the recommended sample size for psychometric analysis of the SWFLS with 646 participants.

Statistical analysis

Following data collection, all statistical analyses in this study were conducted using IBM SPSS Statistics for Windows 26 and IBM Amos 23, and to analyze the McDonald omega (ω), Hayes Macro was used in the SPSS program [43]. Descriptive statistics were used to determine means, standard deviations, and frequencies in the first phase. The SWFLS factor structure was tested using a maximum likelihood estimate CFA, which was conducted after the scale demonstrated normality assumptions. In particular, the original study’s unidimensional factor structure was evaluated by calculating the ratio of the chi-square value to degrees of freedom (df). The criteria for assessing the fit of the unidimensional structure of the SWFLS were as follows: a χ2 / df (p < .05) ratio between 0 and 2 indicated a good fit, while a χ2/df (p < .05) ratio between 2 and 5 indicated an acceptable fit, based on [44]. Along with SRMR and RMSEA values, other goodness of fit indices included GFI, CFI, and NFI values [45]. suggested that GFI, CFI, and NFI values between 0.90 and 0.95, RMSEA values around 0.06, and SRMR values around 0.05 indicate a good model fit. Fornell and Larcker [46] recommend an AVE value larger than 0.50 to achieve convergent validity. Second, the measurement invariance of the SWFLS across genders was assessed using multigroup CFA. Measurement invariance was tested via configural, metric, and scalar invariances, which are the three essential parts of representing measurement equivalence. The goal of configural invariance is to determine whether the factor structure of each group is the same. Furthermore, metric invariance assesses whether there is an invariant relationship between each observed variable and its matching latent attributes. Finally, all factor loadings and intercepts must be invariable to maintain scalar invariance [47, 48]. The study met several criteria suggested by Cheung and Rensvold [49] and Chen [50] to verify measurement invariance: (a) the difference in CFI (ΔCFI) was less than.01; (b) the difference in RMSEA (ΔRMSEA) was less than.015; (c) the difference in SRMR (ΔSRMR) was less than.03; and (d) the difference in χ2 difference value (Δχ2) was not statistically significant (p > .05). Finally, correlation and simple linear regression analyses were conducted to evaluate the convergent validity of the scale. Calculating Cronbach’s alpha (α), composite reliability, McDonald omega (ω), and the intraclass correlation coefficient model 2.1 (ICC2,1) for test-retest reliability was a part of the reliability assessment process.

Results

Preliminary analysis

After screening the missing SWFLS data, we found that none was missing. The normal distribution of the variables was assessed using univariate and multivariate normality distributions. The skewness and kurtosis values for each variable were examined. To identify outliers, z-scores were computed and histograms and box-plot visuals were analyzed. After excluding nine outliers, these values fell within the recommended limits (skewness and kurtosis outside the acceptable range of ± 2), as outlined by Koh [51], ensuring that the criteria for univariate normality were met. The mean scores of the items ranged from M = 4.46 (item 5) to M = 5.48 (item 3). Table 2 presents the descriptive statistics for the SWFLS items.

Construct validity

The preliminary results of the CFA were as follows: χ2 = 41.22, df = 5, χ2/df = 8.24, GFI = 0.975, CFI = 0.986, NFI = 0.984, SRMR = 0.02 and RMSEA = 0.106. In reviewing the proposed indices of change, it was noted that the error correlations between items 1 and 4 were particularly high. The elevated χ2/df was attributed to the interdependence of errors within the indicators. After implementing these recommendations and repeating the analysis, there was a noticeable improvement in the goodness of fit of the model. For the second proposed model, the results were as follows: χ2 = 18.045, df = 4, χ2/df = 4.5, GFI = 0.989, CFI = 0.995, NFI = 0.993, SRMR = 0.013 and RMSEA = 0.074. The factor loadings of the FWLS scale were statistically significant and ranged from 0.79 to 0.92 (see Table 2). In addition, the AVE value of the scale was calculated as 0.75, further confirming the appropriateness of the proposed model fit in the context of the validity study. Figure 1 shows the factor loadings for the SWFLS items.

Table 2 Descriptive statistics and factor loadings for the SWFLS
Full size table
Fig. 1
figure 1

The structure model of the SWFLS

Full size image

Measurement invariance across groups

The one-factor structure of the SWFLS was used as a baseline model to test a series of restrictive models with configural invariance, metric invariance, and scalar invariance. Table 3 shows the measurement invariance analysis findings of the SWFLS in gender groups (female and male samples). As for gender groups, CFI = 0.99, SRMR = 0.017, and RMSEA = 0.06, and all factor loadings reached a significant level (p < .001), which indicated that configural invariance was held across gender groups. In summary, the results of the analysis to test the configuration model showed that the SWFLS could be well described by a one-factor structure of the SWLS in gender groups.

In the second model, which assumed equal loadings of items on a latent variable, the factor loads of the models in the gender groups were equalized. In terms of gender, CFI = 0.991, SRMR = 0.017, and RMSEA = 0.056; therefore, the data fit the metric model (M2) well. Furthermore, ΔCFI =-0.001, ΔSRMR = 0.000, and ΔRMSEA = − 0.015 in comparison with M1 indicate that metric invariance was maintained for both the female and male groups. Consequently, there are equal metrics or scale intervals across gender groups.

Based on the metric model (M2), the scalar invariance model (M3) constrained intercepts to be equal across different groups. The scalar invariance model (M3) limited intercepts to being equal among gender groups based on the metric model (M2). With respect to gender groups, CFI = 0.988, SRMR = 0.017 and RMSEA = 0.054, when compared with the metric model (M2), the values ΔCFI =-0.003, ΔSRMR = 0.000 and ΔRMSEA = -0.002. These values implied that the invariance of intercepts did exist across gender groups. In conclusion, SWFLS’s measurement invariance, which encompasses configural, metric, and scalar invariance, is held across all gender groups. Table 3 shows the results of the measurement invariance across gender groups.

Table 3 Test of measurement invariance across gender groups for the SWFLS
Full size table

Reliability analyses

To determine the discriminative power of each SWFLS item and examine its reliability, item-total statistics were calculated. The corrected item-total correlation coefficients for the five items ranged from 0.75 to 0.85, indicating strong discriminative ability. Reliability for the five-item SWFLS was calculated as the internal consistency coefficient of 0.93 for both Cronbach’s α and McDonald’s ω. In addition, composite reliability was assessed to measure internal consistency, yielding a value of 0.94. To assess test-retest reliability, a subset of university students (n = 48) participated in a two-week interval. The ICC2,1 was used to assess the relative test-retest reliability of each item in the SWFLS. Test-retest reliability (ICC2,1) was high for SWFLS was found to be 0.96 [%95CI = 0.93-0.98] and range from ICC2,1 value 0.90 − 0.85 for items of the SWFLS (reported Table 4).

Table 4 Corrected item-total correlations, reliability analyses and inter-rater analysis for the SWFLS
Full size table

Study 2

Method

Participants and Procedure

The convergent validity study involved 555 participants, aged between 18 and 67 years. Among these individuals, 60% (333) were categorized as female, while 40% (222) were categorized as male. The participants had an average age of 36.4 years, with a standard deviation of 9.73. The approach detailed in the initial study was duplicated in the methodology of this inquiry.

Instrument

Warwick-Edinburgh Mental Well-being scale (WEMWBS)

The WEM-WBS Tennant vd [52]. was developed to assess different aspects of positive mental health, including subjective well-being and psychological functioning. The Turkish version of the scale was adapted from Keldal [53]. It consists of 14 items, and responses are scored by summing the scores of each item on a 1–5 Likert scale. Data from 348 participants were used to evaluate the scale’s internal consistency reliability, yielding a Cronbach’s alpha coefficient of 0.89. A test-retest reliability analysis with 124 individuals was performed one week apart, and it yielded a reliability coefficient of 0.83.

Results

Convergent validity

The relationship between these two scales was examined, and a positive and significant correlation was found between the SWFLS and WEMWBS (r = .483, p < .001). The simple linear regression performed to predict the WEMWBS from the SWFLS yielded the following results: R = .483, R2 = 0.233, F = 168.413, p < .001. These results provide significant support for the convergent validity of SWFLS.

Discussion

This study systematically investigated the psychometric properties of the Turkish version of the SWFLS. The SWFLS was developed using a conceptual framework compatible with the well-established Satisfaction with Life Scale (SWLS) introduced by Diener [19]. According to the CFA results, the one-dimensional factor model consisting of five items showed an acceptable fit after modifications were made with AMOS. Modification indices suggested correlating the errors of item 2 (‘In most ways, my family life is close to ideal’) and item 4 (‘So far, I have gotten the important things I want in my family life’). After following this suggestion, the model showed quite an acceptable fit. As noted by [Harrington [54], correlated measurement errors could arise because of the similarity or proximity of meaning between phrases present in both affirmative and negative statements. In addition to the validity study, the AVE value of the SWFLS was also calculated for convergent validity. The unidimensional factor structure of the scale was also confirmed by the original formulation of the SWLS [19], as well as by SWFLS implementations in different languages, including English [8], Peruvian [55], Portuguese [36], and Spanish [34]. These results confirm the unidimensional configuration of the SWFLS.

The reliability of the Turkish version of the SWFLS was assessed using Cronbach’s alpha coefficient and corrected item-total correlations. The corrected item-total correlations for the items in the Turkish version of the SWFLS ranged from 0.75 to 0.85, showing similar correlations to those observed in the English version of the SWFLS [8]. A consistent pattern has been previously documented in studies on adolescents using SWFLS [34, 36]. In general, items with item-total correlations equal to or greater than 0.30 are considered effective in discriminating individuals [56]. With a Cronbach’s α value of 0.93, the reliability tests for the SWFLS also demonstrated strong reliability. Cronbach’s α coefficients ranged from 0.79 to 0.94 in earlier SWFLS investigations [34, 36] with various sample groups.

Additionally, there was a positive correlation between the WEMWBS and SWFLS (r = .483, p < .01). There is currently no previous study on the relationship between the WEMWBS and SWFLS; therefore, it is impossible to compare the findings. In terms of general mental health, research indicates that those who report higher levels of satisfaction with family life should also exhibit higher levels of psychological well-being [57,58,59]. Therefore, these two factors are expected to have a positive association. These findings are in line with the basic understanding that close relationships and family ties significantly contribute to our emotional state. This research at this time advances the field with these findings.

Limitations and Future Research

While the results of the current study demonstrate the SWFLS’s validity and reliability in gauging family life satisfaction, several study limitations must be acknowledged. First, people who were at least 18 years old are included in the example. Future studies should examine the application of this scale in various scenarios with Turkish children and teenagers. Studies on validity and reliability can be carried out with groups of students (from kindergarten through university) to enhance educational policies and achieve goals, such as figuring out the best psychological counseling services. Second, the SWFLS could be evaluated as a limitation in terms of the test-retest reliability of the participant group and period. Future research should examine test-retest reliability with various participant groups and periods, despite the test-retest correlation’s remarkable temporal consistency. Third, the convergent validity of the association between family life satisfaction and mental well-being was investigated. Convergent validity using only one variable was considered a limitation of our study.

Future studies should examine the connections between various qualities, including support from family, resilience, and positive and negative affect. Additionally, utilizing the SWFSL, longitudinal research should be carried out in the Turkish context to determine whether correlations between mental health and family life satisfaction hold over time. The SWFLS is also applicable to cross-cultural research. The SWFLS is also applicable to cross-cultural research. It will be necessary to pinpoint both the typical and distinctive healing components of family life in today’s multicultural culture, where painful life events such as pandemics and natural disasters occur regularly. Finally, it’s possible that using an online survey for participants decreased the quality of responses. However, the study variables were limited, and the purpose of the study was explained to the participants before administering the scales. Additionally, all participant groups read and signed an informed consent form before completing the sociodemographic data form and study instruments.

Conclusion and implications

Despite these limitations, the current study is the first to demonstrate the construct validity and consistency of the Turkish SWFLS, addressing the need for an internationally recognized measure of family life satisfaction. The Turkish SWFLS has proven to be a reliable and credible instrument for assessing family life satisfaction among individuals aged 18 years and older. This version of the SWFLS offers reliability and validity that has the potential to strengthen family oriented research both theoretically and practically, particularly in the domains of young adult and adult populations. The SWFLS stands out for its brief format, simple instructions, and unidimensional factor structure. It is believed that it can be used in Turkish-speaking contexts and will be useful in terms of both practice and research.

Reduced family life satisfaction is associated with emotional and social difficulties [60], which are frequently affected by family stress and coping mechanisms [61]. This significance is clear in research, family therapy, and psychiatric/psychological counseling. Family therapists and counselors can incorporate the SWFLS into family counseling procedures, enabling them to monitor their clients’ development both before and after the interventions, thereby promoting individual and family well-being. Importantly, the measure functions well for research involving individual interviews, groups, and family counseling with couples.