A Bifactor Exploratory Structural Equation Modeling Representation of the Structure of the Decent Work Scale: Evidence from China

Abstract

The present study investigates the psychometric properties of decent work utilizing the bifactor exploratory structural equation modeling (Bifactor-ESEM) approach. Using a sample of 701 Chinese employees who completed the multidimensional Decent Work Scale (DWS: Duffy et al., 2017), this study reveals the superiority of the Bifactor-ESEM representation of DWS compared to alternative representations of the data (ICM-CFA, Bifactor-CFA, and ESEM). Additionally, the results of measurement invariance in the MIMIC framework indicate the DWS is equivalent in various age and job tenure samples of participants. Finally, the results provide evidence for the criterion validity by confirming the importance of accounting for both the G-factor (representing the global level of decent work) and the S-factors (representing the specific level of decent work), which shows that specific types of decent work explained variance in covariates (i.e., work well-being, life well-being, engagement, and turnover intention) over and above the variance already explained by the G-factor.

Keywords

decent work multidimensionality bifactor exploratory structural equation modeling (Bifactor-ESEM)psychology of working theory China

Although decent work plays a significant role in meeting human needs, enhancing self-worth, and fulfilling individual well-being (Blustein et al., 2019; Duffy et al., 2016), it becomes increasingly difficult to obtain, especially for those living in marginalization or poverty (Blustein et al., 2016). For instance, unpredictable circumstances (e.g., the challenges of profound changes in globalization, advances in artificial intelligence technology, and ebbs and flows of the Novel Coronavirus Pneumonia unseen in a century) cast a shadow on the working situation. In light of these volatile circumstances, the International Labor Organization (ILO, 1999) first proposed the notion of decent work to emphasize the necessity of ensuring that employed individuals have access to decent work conditions regardless of individual characteristics. Duffy and colleagues (2016) further defined decent work as the centerpiece of well-being and work fulfillment from the Psychology of Working Theory (PWT; Blustein, 2006; 2013) perspective. They also developed the reliable and valid Decent Work Scale (DWS; Duffy et al., 2017) to evaluate individuals’ perceptions of their work conditions.

To date, many studies conducted under the PWT framework have demonstrated the prominent role of decent work using the DWS (Duffy et al., 2021). As a promising and inclusive theoretical framework, PWT reckons with decent work as a basic human right vital for individuals' lives (Blustein et al., 2019; Duffy et al., 2016), and interprets the contextual predictors and outcomes for decent work (Duffy et al., 2019). Proliferated empirical studies on decent work demonstrate that the structural and psychological context of work and non-work (i.e., economic constraints, marginalization experiences, work volition, and career adaptability) are crucial antecedents of decent work (Blustein et al., 2019; Duffy et al., 2019; Smith et al., 2020). Moreover, decent work positively affects individuals’ well-being (i.e., job and life satisfaction; Autin et al., 2019; Duffy et al., 2017; Masdonati et al., 2021), work meaning (Di Fabio & Kenny, 2019; Ribeiro et al., 2019), engagement (Ferreira et al., 2019), positive affect (Duffy et al., 2019), and affective commitment (Huang et al., 2021). However, decent work is negatively associated with individuals’ turnover intention (Buyukgoze-Kavas & Autin, 2019; Vignoli et al., 2020), burnout (Ferreira et al., 2019), depression (Nam & Kim, 2019), and negative affect (Duffy et al., 2019). On the other hand, the DWS changes across time following the PWT, which focuses on an individual’s experience during their lifespan. Some researchers discovered that the construct changes over time and provides insight into the connections between predictors (Duffy et al., 2020) and meaningful work and decent work. In general, these studies have contributed to the literature on PWT and have shown that the DWS is reliable and valid.

However, because organizational phenomena are complex, researchers usually use the total score and five subscale scores when investigating the predictors and outcomes associated with decent work in the PWT (e.g., Ribeiro et al., 2019; Wang et al., 2019). In this context, emphasizing the importance of the entire scale and subscales, as well as their relative strength, is beneficial for future decent work studies (Giordano et al., 2020). Morin, Arens, and Marsh (2016) developed the bifactor exploratory structural equation modeling (Bifactor-ESEM) as a framework for accounting for psychometric multidimensionality (e.g., global level of decent work and specific level of decent work) and accurately defining complex psychological characteristics. Accordingly, the current study aims to provide an improved representation of the structure of the DWS while relying on the Bifactor-ESEM framework. Additionally, we use measurement invariance tests to verify the extent to which it generalizes through age and job tenure to similar samples of participants. In a further test of the criterion-validity of the DWS factors, we assess the relations between DWS factors (both G-factor and S-factors) and work well-being, life well-being, engagement, and turnover intention.

Decent Work in China

In the Chinese context, contextual factors have been proven to significantly impact the attainment of decent work and well-being (Chen et al., 2020; Wang et al., 2019). The labor market has transitioned toward a free-market model of hiring and firing with reforms over the past 40 years, resulting in fierce workplace competition and instability. In this context, the household registration system (Hukou, in Chinese) classifies citizens as rural or urban residents, affecting their rights and access to some resources (Wang et al., 2019). Rural migrant workers face insurmountable barriers to equal opportunity in employment and social security provision. They may encounter labor market discrimination and are mostly engaged in informal employment that pays low wages, lacks security and protection, and is not accompanied by a formal employment contract (Cooke et al., 2019). Moreover, with the fierce workplace competition and financial pressure, Chinese employees may have excessive overtime. China has emerged with a “996” work culture, referring to working 12 hours a day (9 a.m.–9 p.m.) and 6 days a week (Bloomberg, 2019). Compared to other nations, Chinese workers work longer hours and have inadequate free time and rest (Nie et al., 2015). Furthermore, Guanxi refers to interpersonal connections and social networks, which significantly influences the potential resource that facilitates job promotion and career success (Chen et al., 2013). The relevance of cross-cultural concepts shows that, according to PWT, contextual factors affect career choices. Thus, it is crucial to provide more accurate evidence for decent work in the Chinese context.

Measurement of the Decent Work

The DWS, which is widely recognized in the PWT literature (Duffy et al., 2019; Wang et al., 2019), consists of 15 items and five dimensions: physically and interpersonally safe working conditions; access to healthcare; adequate compensation; hours that allow for free time and rest; and organizational values that complement family and social values (Duffy et al., 2017). The instrument has been used with diverse samples, such as sexual minorities (England et al., 2020; Tebbe et al., 2019), minoritized racial/ethnic groups (Douglass et al., 2017; Duffy et al., 2019), and colleges (Kim et al., 2019). Moreover, the DWS is also commonly accepted in various countries and cultural settings, including the United States (Duffy et al., 2019; Duffy et al., 2020), Australia (McIlveen et al., 2021), China (Wang et al., 2019), South Korea (Kim et al., 2019), and Sub-Saharan Africa (Atitsogbe et al., 2021). Accordingly, this instrument has received psychometric evidence for its reliability and predictive, convergent and discriminant validity.

Despite the robust psychometric properties reported in previous studies, there remain limitations on the validity of the DWS (Duffy et al., 2017; Dodd et al., 2019). The limitations are as follows: (a) CFA, the widely accepted approach to assessing the structure of psychological constructs, is increasingly questioned due to independent cluster assumption (ICM) (Su et al., 2019). Marsh et al. (2014) argued that this assumption is impractical and excessively strict for instruments measuring complex multidimensional dimensions. The ICM-CFA model is highly unidimensional, which means each item can only reflect on a latent factor. It cannot explain a common but important issue in multidimensional variable domains that two sources and their relative strength of construct-relevant psychometric multidimensionality (Giordano et al., 2020). Even the model fit is satisfactory, the factor correlations may be inflated (Asparouhov et al., 2015; Morin, Arens, Tran, & Caci, 2016). (b) In contrast, the Bifactor model, practical and acceptable in vocational domains, makes it easy to address these questions (Giordano et al., 2020). It provides a flexible alternative model (Morin, Arens, & Marsh, 2016), which allows the total item covariance matrix to be directly separated into one global factor and a series of specific factors, but not only explained by one global factor. Several studies have investigated the psychometric properties of decent work utilizing the Bifactor-CFA model (Duffy et al., 2017; Di Fabio & Kenny, 2019; Nam & Kim, 2019), which discovered that both the general factor and specific factors of decent work are present concurrently. Despite these promising results, Bifactor-CFA may report inaccurate estimates because it neglects the possibility that items may have cross-loadings on non-target specific factors, inflating the general factor’s variance (Morin, Arens, & Marsh, 2016). The CFA and a Bifactor-CFA model are presented on the left side of Figure 1. Therefore, a look into the construct-multidimensionality of decent work using more precise and scientific methods is warranted to provide an accurate conclusion.

Figure 1.

Graphical representation of the alternative models. Note. ICM-CFA = Independent cluster confirmatory factor analysis; ESEM = exploratory structural equation modeling; Bifactor-CFA = bifactor confirmatory factor analysis; Bifactor-ESEM = bifactor exploratory structural equation modeling; D1–D15 = Items; G = Global decent work factor; S1-S5 = Specific decent work factors, S1 = Safe working conditions, S2 = Access to adequate health care, S3 = Adequate compensation, S4 = Free time and rest, S5 = Complementary values.

The Bifactor Exploratory Structural Equation Modeling (Bifactor-ESEM) Framework

Although exploratory factor analysis (EFA) provides more precise estimates of latent factor correlations than CFA and is considered necessary for construct-relevant multidimensionality, it has frequently been questioned for being data-driven and unsuitable for confirmatory studies (Howard et al., 2018). In contrast, CFA relies on models that are assumed to be theory-driven. Asparouhov and Muthén (2009) established Exploratory Structural Equation Modeling (ESEM) by combining the advantages of each technique and incorporating them into an overarching structural equation modeling (SEM) framework. This model provides an appropriate model to investigate the source of multidimensionality due to the associations between indicators and the non-target construct in the form of cross-loadings (Su et al., 2019). ESEM yields better discriminant validity by providing more accurate and lower estimates of factor correlations (Asparouhov & Muthén, 2009; Morin, Arens, & Marsh, 2016). The ESEM solution, however, is still limited since it ignores the latent hierarchically higher-order factor, which may inflate cross-loadings (Morin et al., 2016). In practice, both generality and specificity are affirmed in the structure of decent work (Duffy et al., 2017). That is, individuals not only have a gestalt of perceptions of decent work conditions but also perceptions of five specific components of decent work conditions. For instance, a university professor may highly rate their work conditions as decent, yet, they may perceive having inadequate free time and time for rest due to various tasks. According to PWT, both global and specific perceptions of decent work are crucial for the fulfillment of essential human needs and individual well-being (Blustein et al., 2019). Thus, it might not be feasible to use ESEM to look into the psychometric features of decent work.

Morin, Arens, and Marsh (2016) further proposed the comprehensive and flexible Bifactor-ESEM approach, which integrates the Bifactor and ESEM models into an analytical framework to account for both relevant sources of psychometric multidimensionality. The ESEM and Bifactor-ESEM model are shown on the right side of Figure 1. The Bifactor-ESEM model, in theory, overcomes the shortcomings of the ICM-CFA, Bifactor-CFA, and ESEM models, which can better interpret complex psychological characteristics. Given these advantages, researchers have regarded this approach as a promising alternative to describing complex dimensional structures (Morin, Arens, Tran, & Caci, 2016; Howard et al., 2018). Several researchers have used this method to investigate multidimensional constructs, such as the short version of the workplace affective commitment scale (Perreira et al., 2018), the basic psychological needs at work scale (Sánchez-Oliva et al., 2017), and the engaged teacher scale (Perera et al., 2018). In summary, the Bifactor-ESEM model is recommended for verifying construct-relevant psychometric multidimensionality.

The Present Study

First, this research provides a substantive illustration of the sources of construct-relevant psychometric multidimensionality using Chinese working adults’ responses on the Decent Work Scale. Several alternative measurement models (ICM-CFA, Bifactor-CFA, ESEM, and Bifactor-ESEM) were compared initially to achieve this objective.

Second, although recent research revealed the complex dimensional structures through the Bifactor-ESEM solution, they didn’t conduct measurement invariance tests (Howard et al., 2018; Tóth-Király et al., 2019) or only focused on the gender variable when validating the generality of scale (Morin, Arens, & Marsh, 2016; Sánchez-Oliva et al., 2017). Indeed, it is crucial to demonstrate whether the representation generalizes to similar samples of participants like age or tenure (Sánchez-Oliva et al., 2017; Tóth-Király et al., 2018). In practice, the older workers may experience discrimination due to stereotypes, thus resulting in higher insecurity about work conditions. And the actual duration of the Medical Period and annual paid leave depends on job tenure, according to Chinese Labor Law. On the other hand, empirical research also found that employees’ perceptions about decent work may be influenced by their age and tenure. A meta-analysis (Ng et al., 2005) discovered a positive correlation between age, job tenure, and salary separately. Duffy et al. (2017) also observed a positive relationship between tenure and the dimension “Access to healthcare.” Thus, we investigated whether the DWS measurement is equivalent in age and tenure in the multiple-indicator-multiple-causes (MIMIC) framework through three alternative invariance conditions.

Third, to assess the criterion-related validity of the DWS, the present study adopted the optimal model emerging from the first phase of analysis to examine the relations between the relevant constructs and work well-being, life well-being, engagement, and turnover intention. These outcome variables are drawn from theory and practice. They are essential elements in interpreting the PWT and have been demonstrated positively linked to decent work (Autin et al., 2019; Huang et al., 2021). These variables can assist managers in grasping the importance of decent work and the cause of individuals' well-being, engagement, and turnover intention in practice, on the other hand. To explore additive effects, we systematically compared the added value from the model with only the G-factor and the complete model (including G-factor and S-factor) using the ESEM-within-CFA (EWC; Morin et al., 2013; Morin, Arens, & Marsh, 2016) method.

Method

Participants and Procedure

This study consists of 701 working adults in China, including 360 males (51.4%) and 341 females (48.6%), with a mean age of 30.32 years (SD = 7.13; ranging from 19 to 64 years). Education level is categorized as undergraduate degree (n = 440, 62.8%) and below (n = 90, 12.8%), and master degree (n = 165, 23.5%) or above (n = 6, 0.9%). The average job tenure of participants is 8.11 years (SD = 7.57), and the average weekly hours is 45.54 hours (SD = 9.93). Participants were employed in various occupations. For example, firm employees (n = 278, 39.6%), health care workers (n = 175, 25.0%), educators (n = 87, 12.5%), scientific researchers (n = 14, 2.0%), civil servants (n = 19, 2.7%), as well as others (n = 127, 18.1%).

The survey was created and distributed through an online survey website (http://www.obhrm.com/) in China. Following the American Psychological Association Ethics Code Principles and Standards (2017, Standard 9), we delivered an online informed consent on the first page of the survey, which stated the voluntary nature of participation and assurance of confidentiality and anonymity. The questionnaires also include the decent work scale, the scale of related constructs, and demographic information such as gender, age, job tenure, and weekly work hours. In the present study, 851 participants filled out the questionnaires. A total of 701 responses (82.37% response rate) were retained for analysis after eliminating fast responses (response time faster than 2 seconds per item, Huang et al., 2012), pattern answers (e.g., 111111, 555555, DeSimone et al., 2015), and outliers using the Mahalanobis distance (Meade & Craig, 2012). There were no missing responses due to the online questionnaire design. Moreover, the absolute values of skewness and kurtosis under two in the data met the criteria for univariate normality (skewness <3 and kurtosis <10; Chou & Bentler, 1995; Kline, 2005), suggesting non-normality of the data is not considered to be problematic.

Measures

We used Brislin’s (1980) translation and back-translation approach to translate the Decent Work Scale (DWS; Duffy et al., 2017) into Chinese. All items were independently translated into Chinese by two bilingual authors and reached a consensus on the translation of the scale. And then, all authors discussed the discrepancies and inconsistencies between the back-translated version and the original version, following the back translation of the Chinese version of DWS by the third author. Finally, we confirmed the definitive Chinese version of DWS. Additionally, all instruments are graded by a 5-point Likert-type scale (1 = strongly disagree; 5 = strongly agree).

Decent work

Decent work was measured using the DWS (Duffy et al., 2017), which consists of 15 items assessing five components (i.e., safe working conditions, adequate compensation, free time and rest, access to health care, and complementary values). The sample item is, “I get good healthcare benefits from my job.” In the original study of scale development, Duffy et al. (2017) revealed acceptable internal consistency reliability (ranged from .82 to .97). In the present study, the internal consistency for the total DWS score was .90. Furthermore, internal consistency reliability is acceptable for the subscales of safe working conditions (α = .75), access to healthcare (α = .84), adequate compensation (α = .88), free time and rest (α = .82), and complementary values (α = .90).

Work well-being

The Work Well-Being subscale of the Employee Well-Being Scale (EWB) developed by Zheng et al. (2015) was used to measure respondents' well-being at their current work. It comprises of six items. A sample item is, “I am satisfied with my work responsibilities”. Zheng et al. (2015) reported good reliability (α = .92) and significant relations with job satisfaction, affective organizational commitment, turnover intention, and job performance among Chinese working adults. In present study, the internal consistency was .89.

Life well-being

We used the Life Well-Being subscale of the EWB (Zheng et al., 2015) to evaluate their perception of life well-being. Which is a six items self-report inventory, a sample item as the following: “Most of the time, I do feel real happiness”. Zheng et al. (2015) reported that the scale’s internal consistency was higher than .90, which was significantly associated with job satisfaction, affective organizational commitment, turnover intention, and job performance among Chinese employees. Cronbach’s alpha was .90 in the present study.

Engagement

The Chinese version of the Utrecht Work Engagement Scale (UWES) was utilized to evaluate participants’ work engagement. Recently, Schaufeli et al. (2019) introduced the shorter version of UWES, which has three items and indicates the extent to which an individual engages in their work. Sample items are as follows: (1) “At my work, I feel bursting with energy” (vigor); (2) “I am enthusiastic about my job” (dedication); (3) “I am immersed in my work” (absorption). Reliability and validity evidence presented in the original study (Schaufeli et al., 2019). According to Xu et al. (2022), the scale had good reliability (higher than .80) and was significantly associated with work need satisfaction and life well-being in the Chinese context. In the present study, Cronbach’s alpha was .85.

Turnover intention

The Chinese version of the Turnover Intention Scale initially developed by Mobley et al. (1978) was adopted to measure participants’ intention to leave from their current organization. It comprises of four items, and a sample item is, “I will search for other alternative job opportunities”. Wang et al. (2019) found the scale has good internal consistency reliability (α = .87) and is significantly associated with work volition and decent work in the Chinese context. The estimated internal consistency of the scale in the current study was .90.

Statistical Analyses

Model estimation

In the first phase, we estimated and contrasted ICM-CFA, Bifactor-CFA, ESEM, and Bifactor-ESEM representations of participants’ responses to the DWS. In the ICM-CFA model, items were only allowed to load onto the priority factors of decent work, and the correlation between the factors was freely estimated, but the cross-loadings on other factors were not allowed. In the Bifactor CFA model, items were specified to simultaneously load onto one general factor and five specific factors, and G-factor and S-factor were specified as orthogonal, meaning they did not allow correlations with one another (Morin, Arens, & Marsh, 2016). Notedly, this orthogonality is a prerequisite to the proper disaggregation of the covariance into the G-factor and the S-factors components (Giordano et al., 2020; Reise, 2012). In the ESEM model, all cross-loadings were also freely estimated like the ICM-CFA model while “targeted” to be close to zero using target rotation (Morin, Arens, & Marsh, 2016). Finally, the Bifactor-ESEM model was estimated under classic bifactor assumptions, and all items were allowed to concurrently define a G-factor as well as the five S-factors in a confirmatory manner using an orthogonal bifactor target rotation procedure. We then reported omega coefficients of composite reliability estimated from the alternative models, calculated as ω = (Σ|λ_i|)²/([Σ|λ_i|]² + Σδ_ii) where λ_i stands for the factor loadings and δ_ii corresponds to the error variances (McDonald, 1970).

Measurement invariance

In the second stage of the analysis, we applied the measurement invariance tests to investigate age and tenure differences in responses to the DWS based on the most appropriate model. The multiple-group approach provides full flexibility in assessing the measurement invariance of the optimal measurement model but might not be practical for continuous variables, multiple contrast variables, and their interactions (Morin, Arens, & Marsh, 2016). If the continuous variables (e.g., age) are divided into broad categories, there will be a significant loss of information (Marsh et al., 2013). In contrast, the MIMIC approach provides an appropriate and parsimonious approach to evaluating measurement invariance (Marsh et al., 2013; Morin, Arens, & Marsh, 2016), which permits both categorical and continuous individual difference variables, as well as their interactions. To prevent suboptimal categories in age and job tenure, the present paper adopted the MIMIC approach to evaluate measurement invariance.

By comparing three nested MIMIC models, potential non-invariance of item intercepts was assessed, that is, differential item functioning (Monotonic DIF). In the first (null effect) model, the predictor’s influence on latent means and items’ intercepts were constrained to be zero; in the second (saturated) model, the predictor’s influence on items’ intercepts was freely estimated, but not the latent means. The third (invariant) model assumes the invariance of items’ intercepts across levels of the predictors, which were allowed to influence all latent means but not items’ intercepts. We verified whether the predictors have an effect by comparing the second, third, and first models. Then, the comparison of the second and third models tested whether the effects of the predictors on the items were fully explained by their effects on the latent means, or demonstrated Monotonic DIF. Based on this, the MIMIC approach was adopted to assess the measurement invariance of DWS.

Criterion validity

To provide evidence for the validity of the DWS, we assessed the criterion-related validity of the various decent work factors in the third phase of the analysis. More precisely, we assessed two models: (a) In the first model, both the G-factor (representing the overall level of employees’ decent work) and the S-factors (representing the specific level of employees’ decent work) were allowed to predict all criterion-related variables while simultaneously accounting for the impact of G-factor and S-factors. (b) The second model, consistent with previous research (Howard et al., 2018; Tóth-Király et al., 2019), exclusively explored the G-factor predictive effect using the ESEM-within-CFA (EWC) approach (Morin et al., 2013; Morin, Arens, & Marsh, 2016), which do not investigate the S-factors' predictive effect. And then, we compared the information: the indices of the goodness of fit, standardized regression coefficients (β), and the percentage of explained variance (R²) in the covariates.

Goodness-of-fit assessment

We conducted all statistical analyses in Mplus 8.5 (Muthén & Muthén, 1998–2017). Items with five categories were treated as continuous (Rhemtulla et al., 2012; Sass et al., 2014). Models were assessed by the following fit indices: the χ² (χ² with a significant p value), comparative fit index (CFI), Tucker-Lewis index (TLI), root mean square error of approximation (RMSEA), and standardized root mean square residual (SRMR). According to typical interpretation guidelines (Hu & Bentler, 1999), values greater than .90 and .95 for the CFI and TLI are respectively considered to indicate an adequate and excellent fit to the data, whereas values smaller than .08 or .06 for the RMSEA and SRMR, respectively support an acceptable and excellent fit. Notedly, in line with previous research (Morin et al., 2016; Perera, 2016; Tóth-Király et al., 2018), nested models in MIMIC tests were compared by considering changes (Δ) in goodness-of-fit indices, just like the multiple-group approach. A decrease in the CFI and an increase in RMSEA of less than .01 and .015, respectively, are suggestive of support for a more restrictive model (Chen, 2007; Cheung & Rensvold, 2002).

Results

Factor Structure and Reliability

Table 1 reports the goodness-of-fit indices of the alternative measurement models. The ICM-CFA model achieves an acceptable level of fit to the data based on all goodness-of-fit indices (CFI and TLI ≥.950; RMSEA and SRMR ≤.060). In contrast, the Bifactor-CFA, ESEM, and Bifactor-ESEM solutions provide an excellent fit to the data (∆CFI and ∆TLI > .01; ∆RMSEA > .015) and non-overlapping the 90% RMSEA confidence intervals in comparison with the ICM-CFA models. In addition, The ESEM solution achieves an approximately satisfactory level of fit to the data but only slightly improvements than Bifactor-CFA (∆CLI = +.004, ∆TLI = +.001, ∆RMSEA = −.002). Finally, although the Bifactor-ESEM solution displays an excellent representation of the data according to the CFI and TLI (≥.990), RMSEA and SRMR (≤.020), it displays only marginal improvement relative to the Bifactor-CFA (∆CLI = +.007, ∆TLI = +.006, ∆RMSEA = −.011), and ESEM (∆CLI = +.003, ∆TLI = +.005, ∆RMSEA = −.009), but resulted in a nonsignificant chi square value. Considering the global and general perceptions of decent work by individuals in work life and the practice of Howard et al. (2018), the Bifactor-ESEM model provides an exact fit to the data (Howard et al., 2018).

Table 1.

Goodness-of-Fit Statistics for the Estimated Models.

Model	χ²	df	CFI	TLI	RMSEA (90%CI)	SRMR
ICM-CFA	235.00***	80	.973	.965	.053 (.045, .060)	.041
Bifactor-CFA	122.36***	75	.992	.989	.030 (.020, .039)	.023
ESEM	61.33	40	.996	.990	.028 (.012, .041)	.010
Bifactor-ESEM	37.69	30	.999	.995	.019 (.000, .036)	.007

Note: N = 701. ICM-CFA = independent cluster model-confirmatory factor analysis; Bifactor-CFA = bifactor confirmatory factor analysis; ESEM = exploratory structural equation modeling; Bifactor-ESEM = bifactor exploratory structural equation modeling; χ² = robust chi-square statistic test of exact fit; df = degrees of freedom; CFI = comparative fit index; TLI = Tucker-Lewis index; RMSEA = root mean square error of approximation; 90% CI = 90% confidence interval; SRMR = standardized root mean square residual.

*p < .05; ** p < .01; *** p < .001.

Further, we also conducted a detail parameter estimates associated with these models to enrich the interpretation of the DWS structure by considering the theory expectations (Marsh et al., 2014; Morin, Arens, & Marsh, 2016). Parameter estimates for ICM-CFA and ESEM are reported in Table 2 (correlations) and Table 3 (factor loadings, cross-loadings, and uniquenesses). The five factors in ICM-CFA are well defined by high and significant factor loadings (|λ| = .69 to .92; M = .81; ps < .01), resulting in satisfactory model-based coefficients of composite reliability (ω = .76 to .90, M = .84) for all factors. Consistent with these results, the ESEM model identify the five significant factors of decent work by satisfactory factor loadings (|λ| = .48 to .98; M = .79; ps < .01), and acceptable composite reliability (ω = .72 to.89, M = .83). More importantly, Marsh et al. (2009) recommend that ESEM models should be retained when factor correlations are appreciably reduced relative to CFA correlations. The observation of much comparatively low factor correlations in ESEM (r = .28 to .57; M = .47; ps < .01) compared to CFA (r = .38 to .65; M = .49; ps < .01), which emphasize the importance of the ESEM model when assess the construct-relevant psychometric multidimensionality.

Table 2.

Standardized Factor Correlations for the ICM-CFA (Below the Diagonal) And Five Factor ESEM (Above the Diagonal) Solutions.

	1	2	3	4	5
1. Safe		.49***	.56***	.36***	.55***
2. Health	.52***		.53***	.28***	.47***
3. Comp	.65***	.56***		.45***	.57***
4. Rest	.38***	.27***	.44***		.39***
5. Values	.61***	.48***	.59**	.38***

Note: N = 701. ICM-CFA = independent cluster model-confirmatory factor analysis; ESEM: exploratory structural equation modeling; Safe = Safe working conditions, Health = Access to adequate health care, Comp = Adequate compensation, Rest = Free time and rest, Values = Complementary values. *p < .05; ** p < .01; *** p < .001.

Table 3.

Standardized Factor Loadings (|λ|) and Uniquenesses (δ) for the ICM-CFA and ESEM solutions.

Indicator	ICM-CFA		ESEM
Indicator	\|λ\|	δ	Safe (\|λ\|)	Health (\|λ\|)	Comp (\|λ\|)	Rest (\|λ\|)	Values (\|λ\|)	δ
Safe
D1	.69	.52	.76	.00	−.04	−.02	−.03	.46
D2	.72	.48	.48	−.03	.18	.07	.12	.52
D3	.75	.44	.79	.06	−.05	.00	−.01	.38
ω	.76		.72
Health
D4	.82	.33	.05	.88	−.01	.01	−.04	.36
D5	.89	.21	−.05	.98	−.05	−.01	.01	.14
D6	.72	.48	.02	.61	.10	.00	.06	.50
ω	.85			.87
Comp
D7	.87	.24	.07	.02	.82	.02	−.03	.25
D8	.83	.32	−.13	.03	.90	.01	.01	.26
D9	.84	.30	.11	.02	.71	−.00	.06	.31
ω	.88				.85
Rest
D10	.68	.54	.08	−.02	.16	.60	−.06	.53
D11	.87	.24	−.08	.02	−.05	.95	.01	.18
D12	.81	.34	.04	.00	−.07	.80	.02	.37
ω	.83					.83
Values
D13	.84	.30	.07	.02	.00	.01	.78	.30
D14	.92	.15	−.01	.01	.00	−.01	.92	.15
D15	.83	.31	−.03	−.02	−.01	.00	.86	.30
ω	.90						.89

Note: ICM-CFA = independent cluster model-confirmatory factor analysis; ESEM: exploratory structural equation modeling; ω = omega coefficient of composite reliability; Safe = Safe working conditions, Health = Access to adequate health care, Comp = Adequate compensation, Rest = Free time and rest, Values = Complementary values; λ: Factor loading; δ: Item uniqueness; Target factor loadings are in bold. Non-significant parameters (p ≥ .05) are marked in italics.

Subsequently, we contrasted the Bifactor-ESEM model, and the parameter estimates are reported in Table 4. Aside from the exact fit to the data mentioned before, the Bifactor-ESEM solution reveals five highly reliable S-factors (ω = .44 to.71, M = .57) by strong and positive factor loadings (|λ| = .26 to .84; M = .56; ps < .01). Among them, the S1 factor (Safe working conditions, |λ| = .26 to .60; M = .45) and S3 factor (Adequate compensation, |λ| = .34 to .84; M = .58) appears to be slightly weak defined than other S-factors. Importantly, the Bifactor-ESEM simultaneously revealed the well-defined G-factor (ω = .88) by high and significant factor loadings (|λ| = .37 to .76; M = .57; ps < .01), representing general decent work condition. This implies that the structure of decent work does appear to tap into relevant specificity and add information to the G-factor, which is a crucial advantage beyond the ESEM model (Howard et al., 2018). Further examination of the Bifactor-ESEM finds that some non-target cross-loadings are still significant (|λ| = .00 to .11; M = .04), demonstrating the superiority of this solution over Bifactor-CFA solution. The Bifactor-ESEM solution allows for more precise estimates of the various sources of construct-relevant multidimensionality present in the instrument. In brief, both the exact fit to the data and the theoretical conformity of the parameter estimates support that the Bifactor-ESEM model should be retained.

Table 4.

Standardized Factor Loadings (|λ|) and Uniquenesses (δ) for the Bifactor CFA and ESEM solutions.

Indicator	Bifactor-CFA
Indicator	G \|λ\|	S \|λ\|	δ	G \|λ\|	Safe (\|λ\|)	Health (\|λ\|)	Comp (\|λ\|)	Rest (\|λ\|)	Values (\|λ\|)	δ
Safe
D1	.51	.55	.44	.51	.50	.00	−.05	−.03	−.01	.48
D2	.66	.24	.51	.64	.26	−.06	.03	.04	.05	.51
D3	.58	.53	.39	.55	.60	.06	−.03	−.01	.03	.33
ω		.42			.44
Health
D4	.51	.61	.37	.48	.11	.67	.09	.00	.03	.31
D5	.54	.76	.14	.55	−.03	.71	−.02	−.04	.03	.18
D6	.56	.45	.48	.60	−.07	.43	−.03	−.05	−.03	.45
ω		.64				.64
Comp
D7	.73	.46	.26	.74	.01	.00	.42	.02	−.05	.27
D8	.64	.58	.26	.63	−.06	.06	.64	.04	.03	.19
D9	.74	.38	.31	.76	.00	−.02	.34	−.02	−.02	.30
ω		.47					.46
Rest
D10	.47	.50	.54	.48	.00	−.07	.05	.49	−.07	.52
D11	.39	.82	.17	.37	−.02	.01	.04	.84	.05	.15
D12	.40	.68	.37	.41	.01	−.03	−.04	.67	.02	.38
ω		.71						.71
Values
D13	.64	.54	.30	.63	.06	.02	−.01	.01	.54	.30
D14	.67	.64	.15	.68	−.02	.01	−.03	−.00	.63	.15
D15	.57	.61	.30	.57	.01	.01	.02	.02	.62	.30
ω	.88	.62		.88					.62

Note: Bifactor-CFA = bifactor confirmatory factor analysis; Bifactor-ESEM = bifactor exploratory structural equation modeling; G-Factor = Global decent work factor; S-Factor = Specific decent work factors; ω = omega coefficient of composite reliability; Safe = Safe working conditions, Health = Access to adequate health care, Comp = Adequate compensation, Rest = Free time and rest, Values = Complementary values; λ: Factor loading; δ: Item uniqueness; Target factor loadings are in bold. Non-significant parameters (p ≥ .05) are marked in italics.

Measurement Invariance

Following previous studies Morin, Arens, & Marsh 2016; Perera, 2016), we conducted measurement invariance tests to investigate whether the retained Bifactor-ESEM representation generalizes to similar samples of participants through the MIMIC framework. The results from the MIMIC models are shown in Table 5.

Table 5.

Test of Measurement Invariance of DWS Across Age and Job Tenure.

Model	χ²	df	CFI	TLI	RMSEA (90%CI)	CM	△χ² (△df)	△CFI	△TLI	△RMSEA
Age
M1. MIMIC(Null)	75.42**	45	.995	.986	.018 (.018, .043)	M1	17.72 (6)	+.002	+.004	+.008
M2. MIMIC(Saturated)	57.70*	39	.997	.990	.026 (.009, .040)	M1	37.87 (15)	+.004	+.007	+.001
M3. MIMIC(Invariant)	37.55	30	.999	.995	.019 (.000, .036)	M2	20.15 (9)	+.002	+.005	−.007
Job tenure
M1. MIMIC(Null)	67.14*	45	.996	.990	.026 (.011, .039)	M1	29.45 (15)	+.004	+.005	−.007
M2. MIMIC(Saturated)	37.69	30	.999	.995	.019 (.000, .036)	M1	11.02 (6)	+.001	+.001	−.001
M3. MIMIC(Invariant)	56.12	39	.997	.991	.025 (.000, .036)	M2	18.43 (9)	−.002	−.004	+.006

Note: N = 701. χ² = robust chi-square test of exact fit; df = degrees of freedom; CFI = comparative fit index; TLI = Tucker-Lewis index; RMSEA = root mean square error of approximation; 90% CI = 90% confidence interval of the RMSEA; CM = comparison model; ∆ = change in fit information relative to the CM; *p < .05; ** p < .01; *** p < .001.

First, the model of null effect (M1) was tested by constrained latent means and items’ intercepts of predictors to be zero, providing a satisfactory fit to the data across age and job tenure. The indices for age are: χ² (45) = 75.42, p < .01, CFI = .995, TLI = .986, RMSEA = .018. Indices for job tenure are: χ² (45) = 67.14, p < .05, CFI = .996, TLI = .990, RMSEA = .026. Fit indices of models suggest age and job tenure have limited effects on the factor structure.

Next, the saturated model (M2) and invariant model (M3) were then progressively assessed by constraining the latent means to be zero and constraining the items’ intercepts to be zero. The fit indices do not exceed the recommended guidelines (ΔCFI and ΔTLI <.01, ΔRMSEA <.015) by comparison of the invariant model and null effect model, the invariant model, and null effect model, further supporting that the effects of the predictors (age and job tenure) are not limited to latent means. And then, there is no evidence supporting monotonic DIF (ΔCFI and ΔTLI <.01, ΔRMSEA <.015) through contrasted the saturated model and invariant model. Thus, the Bifactor-ESEM representation of the DWS did not differ significantly between age and job tenure.

Criterion Validity

This research investigated the predictive effects of the decent work G-factor and S-factors on outcomes (i.e., work well-being, life well-being, engagement, and turnover intention). As shown in Table 6, there is a significant positive correlation between the G-factor of decent work and work well-being (62%, explaining 62% of its invariance), life well-being (R² = 61%), engagement (R² = 42%), and a significant negative correlation between the G-factor of decent work and turnover intention (R² = 55%). Based on the multidimensional construct of decent work, we allowed the five factors of decent work to relate to the covariate, and simultaneously the G-factor’s effects are maintained in the complete model. When the S-factors were also considered in the predictive model, the proportion of explained variance increased: (a) from 62 to 63% for work well-being; (b) from 61 to 63% for life well-being; (c) from 42 to 45% for engagement; and (d) from 54.8 to 55.2% for turnover intention. These increases in explained variance are accompanied by increases in goodness-of-fit (complete model: CFI = .998; TLI = .995; RMSEA = .019; restricted model: CFI = .995; TLI = .991; RMSEA = .025).

Table 6.

Relations with Covariates: Standardized Coefficients

Covariates	G-Factor	R ²	Safe	Health	Comp	Rest	Values	G-Factor	R ²
Work well-being	.79***	.62	.17***	−.04	.04	.01	.09*	.77***	.63
Life well-being	.78***	.61	.22***	.02	.05	.04	.03	.76***	.63
Engagement	.65***	.42	.25***	−.02	.03	−.06	.05	.62***	.45
Turnover intention	−.74***	.55	−.08	.00	−.07*	.02	−.09*	−.73*	.55

Note: G-factor = the global level of decent work; Safe = Safe working conditions, Health = Access to adequate health care, Comp = Adequate compensation, Rest = Free time and rest, Values = Complementary values; *p < .05; ** p < .01; *** p < .001.

As expected, there are some significant relationships between the S-factors of decent work and the covariates. “Safe working conditions” factor positively correlates with work well-being (β = .17, p < .001), life well-being (β = .22, p < .001), and engagement (β = .25, p < .001). “Adequate compensation” factor negatively relates with turnover intention (β = −.07, p < .05) “Complementary values” factor significantly associates with work well-being (β = .09, p < .05), and turnover intention (β = −.09, p < .05). In summary, the results provide evidence for the criterion validity of the DWS by the additive effect.

Discussion

The present study aimed to investigate the psychometric properties of the DWS using the Bifactor-ESEM in the Chinese context. The Bifactor-ESEM framework is the appropriate approach to assess two sources of construct-relevant psychometric multidimensionality in organizational research (Perreira et al., 2018). Based on the Bifactor-ESEM representation, we further conducted the test of measurement invariance to verify whether the DWS is equivalent in various individuals (i.e., age and job tenure). Additionally, we also investigated the criterion validity through testing the relations between the Bifactor-ESEM representation of decent work and outcomes (i.e., work well-being, life well-being, engagement, and turnover intention).

First, this study revealed and verified the multidimensional construct of DWS using the Bifactor-ESEM approach in the Chinese context (Morin et al., 2016). The Bifactor-ESEM solution is more satisfactory, which simultaneously considers the global decent work and the five specific decent work factors in a single model that is not tainted by inflated factor correlations or measurement errors. Concretely, the results, on the one hand, strongly support the proposition that the Bifactor-ESEM solutions are superior to other solutions (ICM-CFA, Bifactor-CFA, and ESEM) (Morin, Arens, & Marsh, 2016; Howard et al., 2018; Perreira et al., 2018). Although the ESEM model provides more precise and lower estimates of factor correlations than ICM-CFA and Bifactor-CFA, the Bifactor-ESEM model, consistent with real-life decent work, shows a marginally better fit to the data but reveals a well-defined G-Factor (ω = 0.88) accompanied by lower cross-loadings. On the other hand, the results from the remained Bifactor-ESEM solution are in line with previous studies (Duffy et al., 2017; Di Fabio & Kenny, 2019; Ferreira et al., 2019), providing a more accurate consequence when defining both sources of psychometric multidimensionality.

Apart from the construct validity, the DWS also reports acceptable reliability. Especially, the subscale score accounts for a substantive amount of variance beyond the G-factor when the composite reliability of a specific factor met the criterion (ω > 0.5; Perreira et al., 2018). The findings show that the specific factors of health care (ω = 0.64), time and rest (ω = 0.71), and complementary values (ω = 0.62) had a substantial amount of specificity of its own, but the specific factors of safe work conditions (ω = 0.44) and adequate compensation (ω = 0.46) are marginally satisfactory. That implies the fact that the structure of decent work taps into meaningful specificity once considering the G-factor (Sánchez-Oliva et al., 2017). And two target loadings of safe work conditions and three target loadings of adequate compensation exceed 0.3, indicating that the two specific factors still have an acceptable level of specificity beyond the G-factor (Gu et al., 2020). Briefly, the DWS has satisfactory reliability and stable multidimensional construct validity in the Chinese context, which extends previous studies (Duffy et al., 2017; Masdonati et al., 2021) and may facilitate cross-cultural empirical research progress on decent work in China.

Second, the results of measurement invariance tests in a MIMIC framework revealed that the DWS was found to generalize to various age and job tenure samples of participants. Previous scholars believed that splitting the variables rely on the mean to create two categorical groups may simplify individual differences and lead to limited conclusions (MacCallum et al., 2002). Accordingly, age and job tenure are classic continuous variables, measurement invariance tests utilized the MIMIC approach in the present study to prevent suboptimal age and job tenure categorization. The MIMIC framework is a more parsimonious and appropriate approach to test measurement invariance when the tested variables are continuous (Morin, Arens, & Marsh, 2016; Tóth-Király et al., 2018). The minimal change in fit indices of the null effect model, the saturated model, and the invariant model indicates that the DWS could be used equally for various age and job tenure of participants.

Third, based on PWT, this research investigated the criterion validity of the Bifactor-ESEM solution, underlining the importance of considering both the G-factor and S-factor with the regression coefficients and higher proportion of explained variance in well-being, engagement, and turnover intention. Results suggest that employees' work well-being, life well-being, engagement, and turnover intention are greatly predicted by their global level of decent work. More precisely, whereas decent work is positively linked with individuals' well-being (i.e., work and life) and engagement, it is negatively related to individuals' turnover intention. This is in line with the assertion of PWT that global decent work is associated with increased well-being and work fulfillment (Blustein et al., 2016). Besides, the results also show that some S-Factors uniquely predict higher levels of well-being (i.e., work and life) and engagement, and lower levels of turnover intention, over and above the predictive effects of employees' global levels of decent work, which also support PWT.

Practical Implications

Although recent studies on decent work have confirmed the reliability and validity of DWS, the conclusion might be inaccurate due to the limitations of the method. The Bifactor-ESEM framework was used in this research to demonstrate the psychometric properties of the DWS in the Chinese context. The current study provides practitioners with a useful instrument for assessing employees' work environments, and proposing improvements. The reliable and valid measurement tool also facilitates vocational research in China and contributes to cross-cultural empirical research progress on PWT. Specifically, future studies could identify the different impacts of the structural and psychological context on both the G-factor and S-factors of decent work using the ESEM-within-CFA. Besides, future studies could consider the Bifactor-ESEM framework to provide more precise evidence on the relevant multidimensional construct (e.g., the Work Volition Scale, Duffy et al., 2012) in PWT.

Moreover, the present study extends the PWT by highlighting the importance and complementary role of global and specific levels of decent work on work outcomes. This implies that individuals may cultivate the overall and specific perception of decent work at the same time. That could be ineffective in the human resource management practice when only concerned with employees' general perceptions of their work and ignored their specific perception (e.g., free time and rest). Hence, to satisfy employees' well-being and keep them, managers and human resource practitioners should also pay attention to individuals' specific perceptions of safe work conditions, health care, adequate compensation, free time and rest, and complementary values.

Limitations and Future Research

There are some limitations in the present study that should be considered when interpreting the results. First, this research used cross-sectional data, which precludes directional or causal inferences. Duffy et al. (2020) also emphasize the importance of capturing change in future longitudinal studies of decent work. To reduce the risk of these issues, more sophisticated longitudinal, experimental, or intervention investigations would be necessary. Additionally, although the present study demonstrates the reliability and validity of the DWS in Chinese working adults, the generalizability of the findings may be limited. Future research not only can replicate the results using samples in other cultural contexts but also use a more flexible approach for evaluating measure invariance such as the moderated nonlinear factor analysis (MNLFA) method (Bauer, 2017). Finally, we only investigate the criterion validity using work well-being, life well-being, engagement, and turnover intention. Future studies could assess the criterion-related validity of the Bifactor-ESEM solution by using various variables, such as mental health (Duffy et al., 2021), work meaning (Dodd et al., 2019), and need satisfaction (Autin et al., 2019).

Footnotes

Declaration of Conflicting Interests

The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The author(s) disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: This work was supported by the National Science Foundation of China (Grant no. 71772171), Public Computing Cloud, Renmin University of China and Outstanding Innovative Talents Cultivation Funded Programs 2021 of Renmin University of China.

ORCID iDs

Doudou Liu

Chaoping Li

Yan Xu

References

American Psychological Association (2017). Ethical principles of psychologists and code of conduct. Amended effective June 1, 2010, and January 1, 2017 https://www.apa.org/ethics/code/

Asparouhov

Muthén

(2009). Exploratory structural equation modeling. Structural Equation Modeling: A Multidisciplinary Journal, 16(3), 397–438. https://doi.org/10.1080/10705510903008204

Asparouhov

Muthén

Morin

A. J. S.

(2015). Bayesian structural equation modeling with cross-loadings and residual covariances. Journal of Management, 41(6), 1561–1577. https://doi.org/10.1177/0149206315591075

Atitsogbe

K. A.

Kossi

E. Y.

Pari

Rossier

(2021). Decent work in Sub-Saharan Africa: An application of psychology of working theory in a sample of Togolese primary school teachers. Journal of Career Assessment, 29(1), 36–53. https://doi.org/10.1177/1069072720928255

Autin

K. L.

Duffy

R. D.

Blustein

D. L.

Gensmer

N. P.

Douglass

R. P.

England

J. W.

Allan

B. A.

(2019). The development and initial validation of need satisfaction scales within the psychology of working theory. Journal of Counseling Psychology, 66(2), 195–209. https://doi.org/10.1037/cou0000323

Bauer

D. J.

(2017). A more general model for testing measurement invariance and differential item functioning. Psychological Methods, 22(3), 507–526. https://doi.org/10.1037/met0000077

Bloomberg (2019). China tech, “996” means work, work and more work. Bloomberg. https://www.bloomberg.com/news/articles/2019-05-11/inchina-tech-996-means-work-work-and-more-work-quicktake

Blustein

D. L.

(2006). The psychology of working: A new perspective for career development, counseling, and public policy. Routledge.

Blustein

D. L.

(2013). The psychology of working: A new perspective for a new era. In Blustein

D. L.

(Ed.), The Oxford handbook of the psychology of working (pp. 3–18). Oxford University Press.

10.

Blustein

D. L.

Kenny

M. E.

Di Fabio

Guichard

(2019). Expanding the impact of the psychology of working: Engaging psychology in the struggle for decent work and human rights. Journal of Career Assessment, 27(1), 3–28. https://doi.org/10.1177/1069072718774002

11.

Blustein

D. L.

Olle

Connors-Kellgren

Diamonti

A. J.

(2016). Decent work: A psychological perspective. Frontiers in Psychology, 7, 407. https://doi.org/10.3389/fpsyg.2016.00407

12.

Brislin

R. W.

(1980). Translation and content analysis of oral and written material. In Triandis

H.C.

Berry

J. W.

(Eds.), Handbook of cross-cultural psychology: Methodology (pp. 389–444). Allyn and Bacon.

13.

Buyukgoze-Kavas

Autin

K. L.

(2019). Decent work in Turkey: Context, conceptualization, and assessment. Journal of Vocational Behavior, 112, 64–76. https://doi.org/10.1016/j.jvb.2019.01.006

14.

Chen

F. F.

(2007). Sensitivity of goodness of fit indexes to lack of measurement invariance. Structural Equation Modeling, 14(3), 464–504. https://doi.org/10.1080/10705510701301834

15.

Chen

S.-C.

Jiang

(2020). Decent work in a transition economy: An empirical study of employees in China. Technological Forecasting and Social Change, 153, 119947. https://doi.org/10.1016/j.techfore.2020.119947

16.

Cheung

G. W.

Rensvold

R. B.

(2002). Evaluating goodness-of-fit indexes for testing measurement invariance. Structural Equation Modeling, 9(2), 233–255. https://doi.org/10.1207/S15328007SEM0902_5

17.

Chou

C. P.

Bentler

P. M.

(1995). Estimates and tests in structural equation modeling. In Hoyle

(Ed.), Structural equation modeling: Concepts, issues, and applications (pp. 36–55). Sage.

18.

Cooke

F. L.

Bian

(2019). The prospect of decent work, decent industrial relations and decent social relations in China: Towards a multi-level and multi-disciplinary approach. The International Journal of Human Resource Management, 30(1), 122–155. https://doi.org/10.1080/09585192.2018.1521461

19.

DeSimone

J. A.

Harms

P. D.

DeSimone

A. J.

(2015). Best practice recommendations for data screening. Journal of Organizational Behavior, 36(2), 171–181. https://doi.org/10.1002/job.1962

20.

Di Fabio

Kenny

M. E.

(2019). Decent work in Italy: Context, conceptualization, and assessment. Journal of Vocational Behavior, 110, 131–143. https://doi.org/10.1016/j.jvb.2018.10.014

21.

Dodd

Hooley

Burke

(2019). Decent work in the UK: Context, conceptualization, and assessment. Journal of Vocational Behavior, 112, 270–281. https://doi.org/10.1016/j.jvb.2019.04.002

22.

Douglass

R. P.

Velez

B. L.

Conlin

S. E.

Duffy

R. D.

England

J. W.

(2017). Examining the psychology of working theory: Decent work among sexual minorities. Journal of Counseling Psychology, 64(5), 550–559. https://doi.org/10.1037/cou0000212

23.

Duffy

R. D.

Allan

B. A.

England

J. W.

Blustein

D. L.

Autin

K. L.

Douglass

R. P.

Ferreira

Santos

E. J. R.

(2017). The development and initial validation of the Decent Work Scale. Journal of Counseling Psychology, 64(2), 206–221. https://doi.org/10.1037/cou0000191

24.

Duffy

R. D.

Blustein

D. L.

Diemer

M. A.

Autin

K. L.

(2016). The psychology of working theory. Journal of Counseling Psychology, 63(2), 127–148. https://doi.org/10.1037/cou0000140

25.

Duffy

R. D.

Diemer

M. A.

Perry

J. C.

Laurenzi

Torrey

C. L.

(2012). The construction and initial validation of the work volition scale. Journal of Vocational Behavior, 80(2), 400–411. https://doi.org/10.1016/j.jvb.2011.04.002

26.

Duffy

R. D.

Gensmer

Allan

B. A.

Kim

H. J.

Douglass

R. P.

England

J. W.

Autin

K. L.

Blustein

D. L.

(2019). Developing, validating, and testing improved measures within the Psychology of Working Theory. Journal of Vocational Behavior, 112, 199–215. https://doi.org/10.1016/j.jvb.2019.02.012

27.

Duffy

R. D.

Kim

H. J.

Allan

B. A.

Prieto

C. G.

(2020). Predictors of decent work across time: Testing propositions from psychology of working theory. Journal of Vocational Behavior, 123, 103507. https://doi.org/10.1016/j.jvb.2020.103507

28.

Duffy

R. D.

Prieto

C. G.

Kim

H. J.

Raque-Bogdan

T. L.

Duffy

N. O.

(2021). Decent work and physical health: A multi-wave investigation. Journal of Vocational Behavior, 127, 103544. https://doi.org/10.1016/j.jvb.2021.103544

29.

England

J. W.

Duffy

R. D.

Gensmer

N. P.

Kim

H. J.

Buyukgoze-Kavas

Larson-Konar

D. M.

(2020). Women attaining decent work: The important role of workplace climate in Psychology of Working Theory. Journal of Counseling Psychology, 67(2), 251–264. https://doi.org/10.1037/cou0000411

30.

Ferreira

J. A.

Haase

R. F.

Santos

E. R.

Rabaça

J. A.

Figueiredo

Hemami

H. G.

Almeida

L. M.

(2019). Decent work in Portugal: Context, conceptualization, and assessment. Journal of Vocational Behavior, 112, 77–91. https://doi.org/10.1016/j.jvb.2019.01.009

31.

Giordano

Ones

D. S.

Waller

N. G.

Stanek

K. C.

(2020). Exploratory bifactor measurement models in vocational behavior research. Journal of Vocational Behavior, 120, 103430. https://doi.org/10.1016/j.jvb.2020.103430

32.

H. L.

Wen

Z. L.

Fan

X. T.

(2020). Investigating the multidimensionality of the work-related flow inventory (WOLF): A bifactor exploratory structural equation modeling framework. Frontiers in Psychology, 740. https://doi.org/10.3389/fpsyg.2020.00740

33.

Howard

J. L.

Gagné

Morin

A. J. S.

Forest

(2018). Using bifactor exploratory structural equation modeling to test for a continuum structure of motivation. Journal of Management, 44(7), 2638–2664. https://doi.org/10.1177/0149206316645653

34.

Huang

J. L.

Curran

P. G.

Keeney

Poposki

E. M.

DeShon

R. P.

(2012). Detecting and deterring insufficient effort responding to surveys. Journal of Business and Psychology, 27(1), 99–114. https://doi.org/10.1007/s10869-011-9231-8

35.

Huang

Shen

Yuan

(2021). How decent work affects affective commitment among Chinese employees: The roles of psychological safety and labor relations climate. Journal of Career Assessment, 30(1), 157–180. https://doi.org/10.1177/10690727211029673

36.

L. T.

Bentler

P. M.

(1999). Cutoff criteria for fit indexes in covariance structure analysis: Conventional criteria versus new alternatives. Structural Equation Modeling: A Multidisciplinary Journal, 6(1), 1–55. https://doi.org/10.705519909540118

37.

International Labor Organization [ILO] (1999). Report of the director-general: Decent work. Proceedings of the International Labor Conference Session (87). International Labor Organization.

38.

Kim

H. J.

Duffy

R. D.

Lee

K. H.

(2019). Application of the psychology of working theory with Korean emerging adults. Journal of Counseling Psychology, 66(6), 701–713. https://doi.org/10.1037/cou0000368

39.

Kline

R. B.

(2005). Principles and practice of structural equation modeling (2nd ed.). Guilford.

40.

MacCallum

R. C.

Zhang

Preacher

K. J.

Rucker

D. D.

(2002). On the practice of dichotomization of quantitative variables. Psychological Methods, 7(1), 19–40. https://doi.org/10.1037/1082-989X.7.1.19

41.

Marsh

H. W.

Lüdtke

Nagengast

Morin

A. J.

Von Davier

(2013). Why item parcels are (almost) never appropriate: Two wrongs do not make a right-camouflaging misspecification with item parcels in CFA. Psychological Methods, 18(3), 257–284. https://doi.org/10.1037/a0032773

42.

Marsh

H. W.

Morin

A. J.

Parker

P. D.

Kaur

(2014). Exploratory structural equation modeling: An integration of the best features of exploratory and confirmatory factor analysis. Annual Review of Clinical Psychology, 10, 85–110. https://doi.org/10.1146/annurev-clinpsy-032813-153700

43.

Marsh

H. W.

Muthén

Asparouhov

Lüdtke

Robitzsch

Morin

A. J. S.

Trautwein

(2009). Exploratory structural equation modeling, integrating CFA and EFA: Application to students’ evaluations of university teaching. Structural Equation Modeling: A Multidisciplinary Journal, 16(3), 439–476. https://doi.org/10.1080/10705510903008220

44.

Masdonati

Massoudi

Blustein

D. L.

Duffy

R. D.

(2021). Moving toward decent work: Application of the psychology of working theory to the School-to-work transition. Advance online publication. https://doi.org/10.1177/0894845321991681

45.

McDonald

R. P.

(1970). The theoretical foundations of principal factor analysis, canonical factor analysis, and alpha factor analysis. British Journal of Mathematical and Statistical Psychology, 23(1), 1–21. https://doi.org/10.1111/j.2044-8317.1970.tb00432.x

46.

McIlveen

Hoare

P. N.

Perera

H. N.

Kossen

Mason

Munday

Alchin

Creed

McDonald

(2021). Decent work’s association with job satisfaction, work engagement, and withdrawal intentions in Australian working adults. Journal of Career Assessment, 29(1), 18–35. https://doi.org/10.1177/1069072720922959

47.

Meade

A. W.

Craig

S. B.

(2012). Identifying careless responses in survey data. Psychological Methods, 17(3), 437–455. https://doi.org/10.1037/a0028085

48.

Mobley

W. H.

Horner

S. O.

Hollingsworth

A. T.

(1978). An evaluation of precursors of hospital employee turnover. Journal of Applied Psychology, 63(4), 408–414. https://doi.org/10.1037/0021-9010.63.4.408

49.

Morin

A. J. S.

Arens

A. K.

Marsh

H. W.

(2016). A bifactor exploratory structural equation modeling framework for the identification of distinct sources of construct-relevant psychometric multidimensionality. Structural Equation Modeling: A Multidisciplinary Journal, 23(1), 116–139. https://doi.org/10.1080/10705511.2014.961800

50.

Morin

A. J.

Arens

A. K.

Tran

Caci

(2016). Exploring sources of construct‐relevant multidimensionality in psychiatric measurement: A tutorial and illustration using the composite scale of Morningness. International Journal of Methods in Psychiatric Research, 25(4), 277–288. https://doi.org/10.1002/mpr.1485

51.

Morin

A. J. S.

Marsh

Nagengast

(2013). Exploratory structural equation modeling: an introduction. In Hancock

G. R.

Mueller

R. O.

(Eds), Structural Equation Modeling: A Second Course (2nd ed., pp. 395–436). IAP.

52.

Muthén

B. O.

Muthén

L. K.

(1998–2017). Mplus user’s guide (8th ed.). Muthén & Muthén.

53.

Nam

J. S.

Kim

S. Y.

(2019). Decent work in South Korea: Context, conceptualization, and assessment. Journal of Vocational Behavior, 115, 103309. https://doi.org/10.1016/j.jvb.2019.05.006

54.

T. W. H.

Eby

L. T.

Sorensen

K. L.

Feldman

D. C.

(2005). Predictors of objective and subjective career success: A meta‐analysis. Personnel Psychology, 58(2), 367–408. https://doi.org/10.1111/j.1744-6570.2005.00515.x

55.

Nie

Otterbach

Sousa-Poza

(2015). Long work hours and health in China. China Economic Review, 33, 212–229. https://doi.org/10.1016/j.chieco.2015.02.004

56.

Perera

H. N.

(2016). Construct validity of the social provisions scale: A bifactor exploratory structural equation modeling approach. Assessment, 23(6), 720–733. https://doi.org/10.1177/1073191115589344

57.

Perera

H. N.

Vosicka

Granziera

McIlveen

(2018). Towards an integrative perspective on the structure of teacher work engagement. Journal of Vocational Behavior, 108, 28–41. https://doi.org/10.1016/j.jvb.2018.05.006

58.

Perreira

T. A.

Morin

A. J. S.

Hebert

Gillet

Houle

S. A.

Berta

(2018). The short form of the workplace Affective commitment multidimensional questionnaire (WACMQ-S): A bifactor-ESEM approach among healthcare professionals. Journal of Vocational Behavior, 106, 62–83. https://doi.org/10.1016/j.jvb.2017.12.004

59.

Reise

S. P.

(2012). The rediscovery of bifactor measurement models. Multivariate Behavioral Research, 47(5), 667–696. https://doi.org/10.1080/00273171.2012.715555

60.

Rhemtulla

Brosseau-Liard

P. É.

Savalei

(2012). When can categorical variables be treated as continuous? A comparison of robust continuous and categorical SEM estimation methods under suboptimal conditions. Psychological Methods, 17(3), 354–373. https://doi.org/10.1037/a0029315

61.

Ribeiro

M. A.

Teixeira

M. A. P.

Ambiel

R. A. M.

(2019). Decent work in Brazil: Context, conceptualization, and assessment. Journal of Vocational Behavior, 112, 229–240. https://doi.org/10.1016/j.jvb.2019.03.006

62.

Sánchez-Oliva

Morin

A. J.

Teixeira

P. J.

Carraça

E. V.

Palmeira

A. L.

Silva

M. N.

(2017). A bifactor exploratory structural equation modeling representation of the structure of the basic psychological needs at work scale. Journal of Vocational Behavior, 98, 173–187. https://doi.org/10.1016/j.jvb.2016.12.001

63.

Sass

D. A.

Schmitt

T. A.

Marsh

H. W.

(2014). Evaluating model fit with ordered categorical data within a measurement invariance framework: A comparison of estimators. Structural Equation Modeling: A Multidisciplinary Journal, 21(1), 167–180. https://doi.org/10.1080/10705511.2014.882658

64.

Schaufeli

W. B.

Shimazu

Hakanen

Salanova

De Witte

(2019). An ultra-short measure for work engagement: The UWES-3 validation across five countries. European Journal of Psychological Assessment, 35(4), 577–591. https://doi.org/10.1027/1015-5759/a000430

65.

Smith

R. W.

Baranik

L. E.

Duffy

R. D.

(2020). Psychological ownership within psychology of working theory: A three-wave study of gender and sexual minority employees. Journal of Vocational Behavior, 118, 103374. https://doi.org/10.1016/j.jvb.2019.103374

66.

Tay

Liao

H. Y.

Zhang

Rounds

(2019). Toward a dimensional model of vocational interests. Journal of Applied Psychology, 104(5), 690–714. https://doi.org/10.1037/apl0000373

67.

Tebbe

E. A.

Allan

B. A.

Bell

H. L.

(2019). Work and well-being in TGNC adults: The moderating effect of workplace protections. Journal of Counseling Psychology, 66(1), 1–13. https://doi.org/10.1037/cou0000308

68.

Tóth-Király

Bőthe

Orosz

Rigó

(2019). A new look on the representation and criterion validity of need fulfillment: Application of the bifactor exploratory structural equation modeling framework. Journal of Happiness Studies, 20, 1609–1626. https://doi.org/10.1007/s10902-018-0015-y

69.

Tóth-Király

Morin

A. J.

Bőthe

Orosz

Rigó

(2018). Investigating the multidimensionality of need fulfillment: A bifactor exploratory structural equation modeling representation. Structural Equation Modeling: A Multidisciplinary Journal, 25(2), 267–286. https://doi.org/10.1080/10705511.2017.1374867

70.

Vignoli

Prudhomme

Terriot

Cohen-Scali

Arnoux-Nicolas

Bernaud

J.-L.

Lallemand

(2020). Decent work in France: Context, conceptualization, and assessment. Journal of Vocational Behavior, 116, 103345. https://doi.org/10.1016/j.jvb.2019.103345

71.

Wang

Jia

Hou

Z.-J.

Zhang

Guo

X.-L.

(2019). A test of psychology of working theory among Chinese urban workers: Examining predictors and outcomes of decent work. Journal of Vocational Behavior, 115, 103325. https://doi.org/10.1016/j.jvb.2019.103325

72.

Wang

Lan

(2022). Validation of work need satisfaction scales among Chinese working adults: a psychology of working theory perspective. Journal of Career Assessment, 30(1), 181–199. https://doi.org/10.1177/10690727211032368

73.

Zheng

Zhu

Zhao

Zhang

(2015). Employee well-being in organizations: Theoretical model, scale development, and cross-cultural validation. Journal of Organizational Behavior, 36(5), 621–644. https://doi.org/10.1002/job.1990