Abstract
Grounded in Self-Determination Theory (SDT) and Vallerand's hierarchical model of intrinsic and extrinsic motivation, this research developed the Work Motivation Inventory (WMI). Two studies were conducted to examine the psychometric properties of this scale. Study 1 (N = 497) utilized Exploratory Factor Analysis (EFA) to identify an initial eight-factor structure as a preliminary exploratory finding. In Study 2 (N = 604), Confirmatory Factor Analysis (CFA) was conducted to further refine the scale and validate its final structure. This process resulted in a 40-item full version and a 24-item short form. The validated WMI possesses a seven-factor hierarchical structure, comprising: Amotivation, External Regulation (as a second-order factor with Material and Social sub-factors), Introjected Regulation, Identified Regulation, Intrinsic Motivation to Know, Intrinsic Motivation to Accomplish, and Intrinsic Motivation to Experience Stimulation. Reliability and validity analyses indicated that both the full and short versions possess high internal consistency and criterion-related validity. The WMI provides a psychometrically sound tool for measuring work motivation, offering practical and theoretical insights for building learning organizations, and understanding the underlying structure of work motivation.
Keywords
Introduction
Work motivation, a central topic in organizational research (Baron, 1991), is defined as a set of internal and external forces that initiate and guide work-related behaviors and influence their form, direction, intensity, and duration (Pinder, 2008). Work motivation has significant and far-reaching implications for a variety of organizational and job-related outcomes, including work engagement (Schaufeli et al., 2006), innovative behavior (Kleysen & Street, 2001), job satisfaction, autonomy support, and organizational commitment (Tremblay et al., 2009). Furthermore, recent empirical evidence highlights its critical role in predicting employee retention and ensuring service quality (Senbetu et al., 2025). Consequently, organizations use motivational strategies such as rewards, praise, and performance-based pay to enhance job performance (Bartol & Durham, 2000). Among these, extrinsic rewards are often considered a highly effective method for improving performance (Gibbons, 1997). However, relying solely on such external incentives is increasingly recognized as insufficient for sustaining long-term engagement. A recent comprehensive evidence review conducted by the Chartered Institute of Personnel and Development (Wietrak et al., 2021) highlights that driving work motivation requires a multifaceted approach beyond financial rewards, incorporating factors such as work meaningfulness and autonomy. Crucially, this review identifies Self-Determination Theory (SDT) as one of the most scientifically robust frameworks for understanding these complex motivational dynamics.
Self-Determination Theory (SDT) is a leading framework for work motivation (Deci & Ryan, 1980, 1985, 2000; Ryan & Deci, 2000). As a macro theory of personality and human functioning across contexts, SDT conceptualizes “self-determination” as the sense of volition or psychological freedom experienced during autonomous behavior (Deci & Ryan, 1985). SDT places motivation on a continuum based on an individual's degree of autonomy. This spectrum ranges from amotivation (lacking intent to act) to extrinsic motivation (acting for tangible outcomes), and finally to intrinsic motivation (acting from personal interest and enjoyment) (Ryan & Deci, 2000). Within this continuum, extrinsic motivation is further differentiated into four types of regulatory processes: external regulation, introjected regulation, identified regulation, and integrated regulation. These types vary in their degree of autonomy and occupy distinct positions along the motivational continuum. Specifically, external and introjected regulations are classified as controlled motivation, whereas identified regulation, integrated regulation, and intrinsic motivation are categorized as autonomous motivation (Van den Broeck et al., 2021).
A variety of self-report instruments, mostly grounded in SDT, have been developed to measure work motivation, including the Blais Inventory of Work Motivation (BIWM; Blais et al., 1993), the Work Preference Inventory (WPI; Amabile et al., 1994), the Work Extrinsic and Intrinsic Motivation Scale (WEIMS; Tremblay et al., 2009), the Motivation at Work Scale (MAWS; Gagné et al., 2010), and the Multidimensional Work Motivation Scale (MWMS; Gagné et al., 2015). More recent work has also sought to improve the practical utility of existing work motivation measures by developing shorter forms, such as the shortened WEIMS (Kotera et al., 2022). While SDT provides a nuanced framework for categorizing different types of extrinsic motivation, it typically treats intrinsic motivation as a unidimensional construct. Accordingly, in most existing scales, intrinsic motivation is measured as a single factor. However, Vallerand's hierarchical model of intrinsic and extrinsic motivation (Vallerand et al., 1992) posits that intrinsic motivation comprises three subtypes: intrinsic motivation to know (engaging in activities out of curiosity or a desire to explore new ideas), intrinsic motivation to accomplish (driven by the pursuit of personal goals or standards of excellence), and intrinsic motivation to experience stimulation (seeking enjoyment or sensory pleasure derived from the activity) (Vallerand, 1997; Vallerand et al., 1992). Empirical work confirmed these three forms have distinct antecedents and outcomes (Carbonneau et al., 2012).
This multidimensional structure has been widely validated in cross-cultural research on academic motivation (e.g., Alivernini & Lucidi, 2008; Can, 2015; Kairys et al., 2017; Koludrović & Ercegovac, 2015; Zhang et al., 2016), but its applicability to work contexts has been less explored. We argue this distinction is not only applicable but increasingly critical. The evolution of modern workplaces into “learning organizations” supports extending this multidimensional model, as “organizational learning” — the ongoing acquisition of new knowledge by organizational members, which enhances job performance and enables organizations to adapt and thrive in dynamic environments — has become a central concept in management (Elbawab, 2024; Kezar, 2005; Tsang, 1997). In modern organizations, work and learning are increasingly inseparable. Employees are no longer viewed as having completed their education upon entering the workforce; rather, they must continuously engage with new knowledge and skills in response to organizational demands. This ongoing engagement reflects not only job performance but also a process of learning—and the motivational force underlying it is akin to academic motivation. Therefore, applying the multidimensional model of intrinsic motivation (e.g., distinguishing the drive “to know” from the drive “to accomplish”) to the workplace is a logical and necessary extension. Doing so can enable organizational leaders to better understand the antecedents and consequences of employees’ positive work behaviors, allowing for shaping tasks and environments that foster intrinsic motivation and support optimal organizational functioning.
Work motivation is not solely an individual psychological process but is significantly shaped by macro-social conditions and cultural values such as collectivism (Vo et al., 2022). Recent studies in the Chinese context, particularly within the public administration domain, have proposed “instrumental regulation” as a distinct motivational dimension bridging autonomous and controlled motivation (e.g., Chen et al., 2018; Xu, 2022; Xu & Chen, 2021). These studies highlight the unique role of job security and steadiness. While acknowledging this important perspective, the present study adopts the source-based classification of extrinsic motivation (Gagné et al., 2015), distinguishing between material and social sources. This approach allows us to capture the same substantive factors (e.g., security, benefits) as “material” oriented external regulation, while simultaneously isolating the distinct effects of social pressures (e.g., acknowledgement, approval), which are equally salient in Chinese organizational culture.
Present Study
The present study develops and validates a new Work Motivation Inventory (WMI) by expanding on SDT. We integrate Vallerand et al.'s (1992) tripartite model of intrinsic motivation, differentiating it into motivation to know, to accomplish, and to experience stimulation. This refined structure is intended to capture the nuanced diversity of employees’ intrinsic drives, thereby offering more targeted insights for organizational management practices. Concurrently, we test the subdivision of external regulation into material and social sources, as suggested by Gagné et al. (2015). Our model retains the remaining SDT constructs (introjected, identified, and integrated regulation) and amotivation. This creates our theoretical foundation: a hypothesized nine-factor model of work motivation (see Figure 1) designed to provide more nuanced insights for organizational practice.
Proposed conceptual model of the work motivation inventory.
We tested our model and scale across two studies. Study 1 involved initial item development and a paper-based survey of employed individuals to explore the scale's factor structure and preliminary psychometrics. Study 2 used an online survey with a broader, larger sample to conduct a confirmatory factor analysis of the structure found in Study 1. This second study also assessed the WMI against additional criterion measures and guided the development of a validated short form.
We tested the following hypotheses:
Study 1
Participants
Participants in Study 1 were employees from a commercial bank who were undergoing professional training. A total of 581 responses were initially collected. Based on the recommendations by Velleman and Welsch (1981) and Meade and Craig (2012), insufficient effort responding (IER) behavior was identified using multiple detection methods, including Mahalanobis distance, bogus items, and instructional manipulation checks. Detailed IER screening procedures are provided in the supplementary materials on OSF. After excluding IER cases, a total of 497 valid responses were retained, resulting in a valid response rate of 85.54%. The IER detection rate was within the commonly reported range of 10–15% in prior studies (Huang et al., 2012; Meade & Craig, 2012). Among the valid participants, 318 were female (64.0%) and 179 were male (36.0%), with a mean age of 24.08 years (SD = 2.39). The sample size of N = 497 was determined to be adequate for Exploratory Factor Analysis. This size maintains a subject-to-item ratio of at least ten to one (the initial scale had 51 items).
It is important to note that Study 1 relied on a relatively homogeneous sample of young employees from a single commercial bank. This sampling choice should be understood in light of the role of Study 1 in the overall scale-development process. Study 1 was designed primarily for initial item screening and exploratory factor analysis under a controlled administration setting, rather than for establishing broad generalizability. The offline training context helped ensure standardized administration and attentive responding, which was important at the early item-refinement stage. At the same time, it should be acknowledged that the age and organizational homogeneity of this sample may limit the extent to which all items, especially those involving accumulated work experience or prior changes in work meaning, can be generalized to more experienced employees. Direct information about banking tenure or years of work experience was not collected in Study 1; therefore, this limitation is explicitly acknowledged in the Discussion. To mitigate this concern, Study 2 recruited a broader online working sample from diverse industries and again applied insufficient-effort-response screening procedures before confirmatory analyses.
Measures
The Work Motivation Inventory
Based on the structure and content of existing work motivation scales grounded in SDT, such as the BIWM, WPI, WEIMS, MAWS, and MWMS, as well as academic motivation scales that incorporate the multidimensional structure of intrinsic motivation, we developed a preliminary version of the Work Motivation Inventory. The items were constructed through translation and adaptation of items from existing tools covering different dimensions, as well as through the creation of new items. Experts in organizational psychology and psychological measurement reviewed the Chinese wording to ensure cultural relevance. The resulting initial version of the scale consisted of 51 items. This scale adopts a seven-point Likert format, asking participants to indicate the degree to which each item reflects their reasons for engaging in their current job, with different items corresponding to different types of motivation. An example item is: “Because this job provides me with security.” A detailed table outlining the sources of all items is provided in Supplemental Material 1. As this was a newly developed measure, its factor structure required empirical exploration to assess alignment with the proposed theoretical model.
Criterion Scales
Work Engagement
Work engagement refers to a positive, fulfilling, work-related state characterized by vigor, dedication, and absorption (Schaufeli et al., 2006). It was measured using the Utrecht Work Engagement Scale (UWES), a seven-point Likert scale consisting of 16 items across three dimensions. In the current study, the scale showed excellent internal consistency (Cronbach's α = .93).
Organizational Citizenship Behavior (OCB)
OCB describes discretionary behaviors that benefit the organization but are not part of formal job requirements (Podsakoff et al., 1990). Prior research suggests a positive correlation between OCB and intrinsic motivation, and a negative correlation with extrinsic motivation (Barbuto & Story, 2011). This study used an eight-item, five-point Likert scale developed by Podsakoff et al. (1990), which yielded good internal consistency (Cronbach's α = .86).
Meaningful Work
Meaningful work is defined as a subjective experience of meaningfulness in one's job. It has been shown to correlate negatively with external motivation and positively with intrinsic motivation (Steger et al., 2012). The Work and Meaning Inventory (WAMI), a 10-item, five-point Likert scale, was used to assess this construct (Cronbach's α = .82 in this study).
Individual Innovative Behavior
This construct refers to individual behaviors aimed at generating, introducing, and applying novel and beneficial ideas in the workplace (Kleysen & Street, 2001). It was measured using the Individual Innovative Behavior Measure (IIBM), which includes 14 items across five dimensions on a five-point Likert scale. The scale demonstrated high internal consistency in this study (Cronbach's α = .94).
Burnout
Burnout is a negative psychological condition associated with work, typically involving emotional exhaustion, lack of enthusiasm, cognitive dysfunction, and emotional impairment (Schaufeli et al., 2002). A lack of motivation is known to exacerbate burnout, especially under high pressure and low support (Fernet et al., 2010). The Burnout Assessment Tool (BAT), consisting of four items rated on a five-point Likert scale, was used to measure this construct (Cronbach's α = .85 in this study).
Stress
Stress was measured using the Stress Mindset Measure (SMM; Crum et al., 2013), which assesses individuals’ beliefs about the impact of stress. It includes two dimensions—positive (e.g., “Experiencing stress enhances my performance and productivity”) and negative (e.g., “Experiencing stress depletes my health and vitality”)—across eight items on a five-point Likert scale. Internal consistency for the two subscales in this study was .78 and .67, respectively.
Procedure
The study protocol was approved by the Institutional Review Board of Beijing Normal University. All participants provided informed consent prior to their participation, in accordance with the approved procedures. Data collection for Study 1 was conducted in March 2024 in collaboration with a large commercial bank headquartered in Beijing, China. The survey was administered offline (paper-and-pencil) during a scheduled professional training program for new management trainees. All trainees present at the session were invited to participate on a voluntary basis. With assistance from the organization's HR department, participants were informed that the survey aimed to understand employee psychological and behavioral preferences to improve workplace conditions. They were instructed to complete the questionnaire within a limited time frame based on their first impressions. To ensure honest responding, participants were assured that their responses would be confidential and would not affect their personal performance evaluations. All responses were anonymized before analysis.
Statistical Analysis
All data analyses in Study 1 were conducted using SPSS 27 and Mplus 8.3 (Muthén & Muthén, 1998–2017). After assessing the suitability of the data for factor analysis, an exploratory factor analysis (EFA) was performed using geomin oblique rotation and robust maximum likelihood estimation (MLR), which are recommended when factors are expected to be correlated and the data may deviate from multivariate normality (Asparouhov & Muthén, 2009; Fabrigar et al., 1999; Li, 2016; Muthén & Muthén, 1998–2017). The number of factors and item-retention decisions were determined through an iterative process that combined empirical criteria with theoretical interpretability. We first inspected solutions around the theoretically expected structure and compared models with different numbers of factors. The scree plot and model fit indices, including CFI, TLI, RMSEA, and SRMR, were used to evaluate whether additional factors meaningfully improved model fit and interpretability. At the item level, we examined primary factor loadings, cross-loadings, and whether each item loaded on a theoretically appropriate factor. As a general guideline, items with primary loadings below .30, substantial cross-loadings, or loading patterns inconsistent with their intended dimension were considered for removal. However, final decisions also considered whether retaining an item was necessary for content coverage. EFA was rerun after each major item-removal step until the solution showed acceptable model fit and a theoretically interpretable structure. Internal consistency (Cronbach's alpha) was computed for each subscale to evaluate reliability. Pearson correlations among subscales and with criterion measures were calculated to assess the scale's criterion-related validity.
To assess the potential impact of common method variance (CMV) arising from the use of self-report measures, we additionally conducted Harman's single-factor test in SPSS (Podsakoff et al., 2003). All items from the Work Motivation Inventory and the criterion scales in Study 1 were entered into an exploratory factor analysis using principal axis factoring with an unrotated solution. 21 factors had eigenvalues above 1, and the first factor accounted for 34.306% of the total variance. Because multiple factors emerged and the first factor did not explain the majority of the covariance among the items, CMV is unlikely to pose a serious threat to the substantive findings of Study 1.
To ensure the reproducibility of our findings, the anonymized data, analysis code (Mplus and SPSS syntax), and supplementary materials for both Study 1 and Study 2 are openly available on a third-party archive. These materials can be accessed for review purposes at the following anonymous link: https://osf.io/emnrs/?view_only=218b158157b240a18446edc215087213.
Results
Factor Analysis
Exploratory factor analysis (EFA) was conducted using Mplus 8.3 to explore the underlying factor structure of the Work Motivation Inventory and to refine the items accordingly. The Kaiser–Meyer–Olkin (KMO) measure of sampling adequacy was 0.967, indicating that the data were highly appropriate for factor analysis. Bartlett's test of sphericity yielded a chi-square value of 23778.755 (df = 1275, N = 497, p < .001), suggesting the presence of sufficient correlations among variables to justify factor extraction.
EFA was conducted using geomin oblique rotation and robust maximum likelihood estimation (MLR). We first examined EFA solutions from five to ten factors. Although a solution close to the theoretically expected structure was examined, the integrated regulation items did not form a stable and independent factor. Instead, the items originally designed to measure integrated regulation showed salient loadings on factors that also reflected identified regulation and intrinsic motivation. For example, in the initial 51-item EFA, Integ1 loaded on two factors (.327 and .315), Integ2 loaded on two factors (.313 and .510), and Integ5 also showed salient loadings on two factors (.307 and .344). Integ3 (.527) and Integ4 (.556) loaded primarily on a factor that was not empirically separable as an integrated-regulation factor. These results suggested that the integrated regulation items were not sufficiently distinct from adjacent autonomous motivation dimensions in this sample. This pattern is consistent with prior research indicating that integrated and identified regulation are often difficult to distinguish empirically (Gagné et al., 2015; Vallerand et al., 1992; Van den Broeck et al., 2021).
Based on these results, most integrated regulation items were removed during the item-refinement process. Specifically, Integ1, Integ3, Integ4, and Integ5 were deleted. One item originally developed under integrated regulation, Integ2 (“Because this job is part of my life ideals”), was retained because its content was theoretically close to identified regulation and represented personally valuing one's work as part of broader life goals. In the final CFA and final scale, this item was treated as an identified regulation item and relabeled as Iden5, whereas the original Iden5 item was removed during EFA because of problematic loading patterns.
Other items with factor loadings below .30, substantial cross-loadings, or theoretically inconsistent loading patterns were considered for removal iteratively. EFA was re-run after each deletion until the majority of items demonstrated acceptable loading patterns. To maintain the theoretical coverage and representativeness of the scale, a few items with borderline loadings were retained. For instance, Amo6 (“There's no particular reason; everyone has to work anyway”) was retained as it captures a passive, fatalistic aspect of amotivation that is distinct from the active meaninglessness expressed in other items. Similarly, Ext_soc4 (“Because this job provides me with more social opportunities”) was retained, despite some conceptual ambiguity (which likely explains its borderline loading), to ensure the “External Regulation-Social” factor captured the important aspect of gaining social network access, which is distinct from seeking direct praise or status. A total of seven items—Integ1, Integ3, Integ4, Integ5, Introj1, the original Iden5, and ImEX3—were deleted during the refinement process. The final version retained 44 items and yielded an eight-factor solution. These factors corresponded to the eight dimensions of work motivation, with the integrated regulation dimension removed. The factor names and loading ranges were as follows: Amotivation (Amo): .284–.909; External Regulation–Material (Ext_mat): .319–.853; External Regulation–Social (Ext_soc): .689–.943; Introjected Regulation (Introj): .271–.941; Identified Regulation (Iden): .298–.849; Intrinsic Motivation–To Know (ImTK): .559–.803; Intrinsic Motivation–To Accomplish (ImAC): .302–.948; Intrinsic Motivation–To Experience Stimulation (ImEX): .425–.710. The model showed a good fit to the data: χ2 (622, N = 497) = 1235.651, p < .001, χ2/df = 1.987, CFI = 0.961, TLI = 0.941, RMSEA = 0.045 (90% CI: 0.041–0.048), SRMR = 0.019. These results provide preliminary support for the structural distinctions proposed in Hypothesis 1a and 1b. Detailed factor loadings are provided in Supplemental Material 2. All subsequent testing and analyses were based on this refined version.
Reliability
Cronbach's alpha coefficients were calculated for each dimension. As shown in Table 1, the alpha values of the newly developed Work Motivation Inventory ranged from .724 to .947, indicating good internal consistency reliability.
Descriptive Statistics, Reliability, and Intercorrelations of Dimensions in Study 1's Work Motivation Scale
Note: All correlation coefficients are significant at p < .05 level. Amo = Amotivation; Ext (mat/soc) = External regulation (material/social); Introj = Introjected regulation; Iden = Identified regulation; ImTK = Intrinsic motivation to know; ImAC = Intrinsic motivation to accomplish; ImEX = Intrinsic motivation to experience stimulation.
Intercorrelations
To test the self-determination continuum assumption (Hypothesis 2), we examined the intercorrelations among the subscales. As Table 1 shows, dimensions theoretically closer to each other on the continuum exhibited stronger positive correlations. Furthermore, autonomous motivation dimensions (Iden, ImTK, ImAC, ImEX) showed significant negative correlations with Amotivation (rs ranging from −.529 to −.610). These patterns provide initial support for Hypothesis 2.
Criterion-Related Validity
Criterion-related validity refers to the degree to which a scale correlates with external criteria, typically assessed using correlation coefficients. Table 2 presents the correlations between each dimension of the Work Motivation Inventory developed in Study 1 and the criterion measures. The relationships between work motivation dimensions and the criterion variables were consistent with theoretical expectations. Specifically, the autonomous motivation dimensions (Iden, ImTK, ImAC, ImEX) were significantly and positively correlated with work engagement (UWES), organizational citizenship behavior (OCB), meaningful work (WAMI), and individual innovative behavior (IIBM), with correlation coefficients ranging from r(497) = .364 to .717 (95% CIs [.285 .757], all ps < .001). This suggests that higher levels of autonomous motivation are associated with greater enthusiasm, a stronger sense of meaning in work, and higher levels of innovation. In addition, autonomous motivation dimensions were significantly negatively correlated with burnout (BAT), r(497) = −.563 to −.474 (95% CIs [−.621 −.403], all ps < .001), indicating that individuals with higher autonomous motivation tend to experience lower levels of burnout. Specifically, supporting Hypothesis 3, autonomous motivation dimensions demonstrated significantly stronger positive associations with positive work outcomes compared to controlled motivation dimensions. For example, correlations between autonomous dimensions and Work Engagement (UWES) ranged from .597 to .717, whereas correlations for external and introjected regulations ranged only from .259 to .400. Similarly, autonomous motivation showed stronger negative correlations with Burnout (BAT) (rs from −.474 to −.563) compared to controlled motivation (rs from −.182 to −.306).
Reliability, Descriptive Statistics, and Correlations of Criterion-Related Measures with Dimensions of the Work Motivation Inventory in Study 1
Note: All correlation coefficients are significant at p < .05 level. Amo = Amotivation; Ext (mat/soc) = External regulation (material/social); Introj = Introjected regulation; Iden = Identified regulation; ImTK = Intrinsic motivation to know; ImAC = Intrinsic motivation to accomplish; ImEX = Intrinsic motivation to experience stimulation. UWES = Utrecht Work Engagement Scale; OCB = Organizational Citizenship Behaviors; WAMI = The Work and Meaning Inventory; IIBM = Individual Innovative Behavior Measure; BAT = Burnout Assessment Tool; SMM (N / P) = Stress Mindset Measure (Negative / Positive).
For the controlled motivation dimensions (Ext_mat, Ext_soc, Introj), the two external regulation dimensions (Ext_mat, Ext_soc) showed modest positive correlations with some positive criteria (e.g., UWES, OCB, and WAMI), r(497) = .147 to .426 (95% CIs [.060 .495], all ps < .001). They were also negatively correlated with burnout (BAT), r(497) = −.305 to −.182 (95% CIs [−.383 −.096], all ps < .001), suggesting that controlled motivation may also be linked to lower stress and burnout. Introjected regulation was also positively correlated with positive criteria such as work engagement, OCB, and meaningful work, r(497) = .298 to .401 (95% CIs [.216 .472], all ps < .001), though the correlations were smaller compared to those of autonomous motivation. This reflects the possibility that some individuals maintain work motivation due to internalized external pressure or identification, though the quality of such motivation may be less optimal.
Furthermore, the scale's criterion-related validity was supported by its differential correlations with stress mindsets. Specifically, autonomous motivation dimensions showed robust correlations with an adaptive stress mindset, being strongly associated with a positive stress mindset (SMM-P) (r[497] = .364 to .386, 95% CIs [.285 .459], all ps < .001) and negatively with a negative stress mindset (SMM-N) (r[497] = −.465 to −.404, 95% CIs [−.531 −.327], all ps < .001). While controlled motivation followed a similar correlational pattern, the magnitude of these associations was consistently weaker. This differential pattern demonstrates that the scale not only correlates with related constructs but also accurately captures the qualitative distinctions between higher-quality autonomous motivation and lower-quality controlled motivation, as theorized by SDT.
Taken together, the results support the criterion-related validity of the revised Work Motivation Inventory developed in Study 1. The observed relationships between motivation dimensions and external variables such as work engagement, OCB, burnout, and meaningful work are theoretically consistent, thereby supporting Hypothesis 3.
Study 2
Study 1 demonstrated that the newly developed Work Motivation Inventory consists of eight dimensions and possesses good reliability and validity. However, to avoid sample-specific bias in the findings, Hinkin (1998) recommends using multiple samples to test the factor structure of newly developed scales. Therefore, in Study 2, we used a separate sample of working adults and collected data via an online crowdsourcing platform to further examine the factor structure of the Work Motivation Inventory and gather additional evidence for its validity.
Participants
Participants in Study 2 were recruited via the online crowdsourcing survey platform Wenjuanxing. Informed consent was obtained from all participants prior to their participation in the study. A broader range of working adults were invited to complete the questionnaire, resulting in a total of 637 responses. Following the recommendations of Velleman and Welsch (1981) and Meade and Craig (2012), we applied several statistical techniques and indices to identify insufficient effort responding (IER), including Mahalanobis distance, average long-string index, bogus items, and instructive manipulation check. Detailed IER screening procedures are provided in the supplementary materials on OSF. After data screening, 604 valid responses were retained, yielding an effective response rate of 94.82%. Among the valid participants, 350 were female (57.9%) and 254 were male (42.1%), with a mean age of 35.11 years (SD = 6.94). In addition to being older than the Study 1 sample, participants in Study 2 had more substantial work experience. Their average years of work experience was 8.58 years (SD = 5.95), with a range from 1 to 42 years. Thus, Study 2 provided an opportunity to examine the refined WMI structure in a sample with broader occupational experience than the initial Study 1 training sample, representing over 20 distinct industries (e.g., manufacturing, technology, education, finance, and healthcare).
Measures
Work Motivation Inventory
The Work Motivation Inventory used in this study comprised the 44 items retained from Study 1. The scale is rated on a seven-point Likert scale and includes eight dimensions: Amotivation (Amo): six items; External Regulation–Material (Ext_mat): five items; External Regulation–Social (Ext_soc): five items; Introjected Regulation (Introj): five items; Identified Regulation (Iden): six items; Intrinsic Motivation–To Know (ImTK): six items; Intrinsic Motivation–To Accomplish (ImAC): six items; Intrinsic Motivation–To Experience Stimulation (ImEX): five items.
Criterion Scales
Perceived Organizational Support (POS)
Perceived organizational support refers to employees’ perceptions of the extent to which the organization values their contributions and cares about their well-being (Eisenberger et al., 1986). POS has been found to positively predict intrinsic motivation, integrated regulation, identified regulation, and introjected regulation, while negatively predicting external regulation and amotivation (Tremblay et al., 2009). POS was measured using the short form of the Survey of Perceived Organizational Support (SPOS). This scale comprises eight items rated on a seven-point Likert scale. In the present study, the internal consistency for this scale was .90.
Organizational Commitment (OC)
Organizational commitment refers to the relative strength of an individual's identification with and involvement in a particular organization (Allen & Meyer, 1990). It is positively predicted by all types of motivation except amotivation, and negatively predicted by amotivation (Tremblay et al., 2009). OC was measured using the Affective and Continuance Commitment Scale (ACS), which assesses the affective, continuance, and normative dimensions of organizational commitment. This scale consists of 24 items rated on a seven-point Likert scale. In the present study, internal consistency for the three subscales ranged from .76 to .77.
Job Satisfaction
Job satisfaction represents an employee's cognitive appraisal of how their work environment or context contributes to their well-being (Gagné et al., 2007). It is expected to be positively predicted by identified regulation and intrinsic motivation (Gagné et al., 2007). Job satisfaction was measured using the Satisfaction with Work Scale (SWWS). This scale consists of five items, and is rated on a seven-point Likert scale. In the present study, the internal consistency for this scale was .76.
Turnover Intention
Turnover intention refers to the likelihood that an employee will leave their job within a certain period (Scott et al., 1999). It has been shown to have significant negative correlations with all types of work motivation (Tremblay et al., 2009). Turnover intention was measured using the Intent to Leave Scale (ILS). This scale consists of four items rated on a five-point Likert scale. In the present study, the internal consistency for this scale was .86.
Basic Psychological Need
Basic psychological needs are fundamental requirements that play a core role in individual psychological development and well-being, serving as crucial drivers for intrinsic motivation and psychological health (Ryan & Deci, 2000). A supportive work environment can satisfy employees’ basic psychological needs, thereby positively influencing their intrinsic motivation (Gagné & Deci, 2005). Need satisfaction was measured using The Basic Need Satisfaction at Work Scale (BNS-W; Deci et al., 2001). This scale consists of 21 items assessing the satisfaction of autonomy, competence, and relatedness needs, rated on a seven-point Likert scale. In the present study, the internal consistency for the three subscales ranged from .77 to .89.
Achievement Motivation
Achievement motivation refers to an individual's propensity to strive for success in situations where standards of excellence can be applied, typically differentiated into the motive to achieve success (hope of success) and the motive to avoid failure (fear of failure) (Lang & Fries, 2006). In this study, we predicted that higher levels of achievement motivation would correspond to higher levels of achievement-related intrinsic motivation. Achievement motivation was measured using the Achievement Motives Scale (AMS; Lang & Fries, 2006). This scale comprises 10 items rated on a seven-point Likert scale, with five items measuring hope for success and five items measuring fear of failure. In the present study, the Cronbach's alpha was .79 for the hope for success subscale and .88 for the fear of failure subscale.
Procedure
The study protocol was approved by the Institutional Review Board of Beijing Normal University. All participants provided informed consent prior to their participation, in accordance with the approved procedures. Data collection for Study 2 was conducted in November 2024 via the online crowdsourcing survey platform Wenjuanxing. The platform's sampling service was utilized to reach a broader and more diverse population of working adults. Specific inclusion criteria were set to ensure data quality: participants were required to be currently employed full-time, aged 18 or above, and residents of Mainland China. The survey link was distributed randomly to registered users within the platform's pool who met these criteria. Following Meade and Craig (2012), participants with an average item response time of less than 2 s were automatically excluded by the system to reduce the likelihood of IER. Early stage screening during data collection helped facilitate more efficient data cleaning. Once the target number of valid responses was reached, data collection was terminated, and the responses were exported for analysis.
Statistical Analysis
Data analyses were performed using SPSS 27 and Mplus 8.3. Confirmatory factor analysis (CFA) was conducted using robust maximum likelihood estimation (MLR) to examine whether the data supported the eight-factor structure identified in Study 1. Model fit indices and factor loadings were reported to assess the adequacy of the structure. Descriptive statistics, internal consistency coefficients, and inter-factor as well as factor-to-criterion correlations were also calculated to examine the reliability and criterion-related validity of the scale.
Similarly, to evaluate whether CMV might bias the results of Study 2, we conducted Harman's single-factor test in SPSS. All Study 2 items were subjected to an exploratory factor analysis using principal axis factoring and an unrotated solution. The analysis yielded 26 factors with eigenvalues greater than 1, and the first factor accounted for 29.959% of the total variance. Consistent with the recommendations by Podsakoff et al. (2003), the presence of multiple factors and the absence of a dominant single factor suggest that CMV is unlikely to be a predominant concern in Study 2.
Results
Factor Analysis
The results of the Confirmatory Factor Analysis (CFA) indicated that the model fit was acceptable: χ2 (874, N = 604) = 2352.288, p < .001, χ2/df = 2.691, CFI = 0.903, TLI = 0.895, RMSEA = 0.053 (90% CI: 0.049–0.057), SRMR = 0.059. Given the unequal number of items across dimensions in the current scale, items Amo4, Iden6, ImTK5, and ImAC6 were removed based on their factor loadings in the CFA, ensuring that each dimension contains five items. A subsequent CFA was conducted on the revised version. The results showed an overall improvement in model fit: χ2 (712, N = 604) = 1976.266, p < .001, χ2/df = 2.776, CFI = 0.910, TLI = 0.901, RMSEA = 0.054 (90% CI: 0.051–0.057), SRMR = 0.058. Therefore, Hypothesis 1a was supported, and the 40-item version was retained as the finalized Work Motivation Inventory. The full 40-item scale is presented in Supplemental Material 3.
Furthermore, we tested the potential hierarchical structure of the Work Motivation Inventory as proposed by Gagné et al. (2015), in which External Regulation (Ext) is modeled as a second-order latent variable encompassing two subdimensions: External Regulation–Material (Ext_mat) and External Regulation–Social (Ext_soc). To verify the rationality of this hierarchical structure, a second-order CFA was conducted using Mplus. The results indicated good model fit: χ2 (717, N = 604) = 2006.847, p < .001, χ2/df = 2.799, CFI = 0.908, TLI = 0.900, RMSEA = 0.055 (90% CI: 0.051–0.059), SRMR = 0.058. The standardized factor loadings of the second-order factor Ext on its subdimensions Ext_mat and Ext_soc were .902 and .853, respectively, indicating strong explanatory power. The factor loadings for items under each dimension ranged from .480 to .871, with an average of .728 (see Table 3 for details). These results support the proposed second-order structure of the Ext factor, suggesting that extrinsic regulation dimension can be further differentiated into material and social components, supporting Hypothesis 1b. Accordingly, this study adopts the second-order factor model of work motivation, consisting of two first-order factors and seven second-order factors, with a total of 40 items. The path diagram of the final model is presented in Figure 2.
Path diagram of the final seven-factor hierarchical structure of the work motivation inventory (WMI). Note: Ovals represent latent factors; rectangles represent bundles of observed items (ranges indicate item labels). The model features a hierarchical structure where External Regulation functions as a second-order factor underlying two first-order dimensions: Material and Social. Double-headed arrows indicate correlations between the highest-order latent factors. Single-headed arrows represent standardized factor loadings. Error terms are omitted for visual clarity.
Confirmatory Factor Analysis Factor Loadings of the Work Motivation Inventory in Study 2
Note: Amo = Amotivation; Ext (mat/soc) = External regulation (material/social); Introj = Introjected regulation; Iden = Identified regulation; ImTK = Intrinsic motivation to know; ImAC = Intrinsic motivation to accomplish; ImEX = Intrinsic motivation to experience stimulation.
The Brief Version of the Work Motivation Inventory
Following the development of the Work Motivation Inventory which consists of 40 items across seven dimensions, a short form was developed to enhance practicality and ease of application in real-world assessment contexts and to meet diverse measurement needs. The development and item selection for the short form were guided by the following criteria: (1) the size of the factor loadings; (2) the impact of item removal on overall model fit indices; (3) representativeness and non-redundancy of item content. Specifically, we first prioritized retaining items with higher standardized factor loadings within their respective dimensions based on CFA results, in order to maintain strong psychometric properties. Next, we applied a stepwise item reduction approach by comparing the model fit indices (χ2/df, CFI, TLI, RMSEA, SRMR) of various item combinations and removing items that contributed less to the overall model fit, thus optimizing the scale structure. Finally, we ensured that the selected items maintained comprehensive coverage of the construct, while also emphasizing concise item wording and applicability across different assessment contexts, avoiding redundancy.
After optimization, three items were retained for each dimension, resulting in a 24-item short form of the Work Motivation Inventory. The items constituting the short form are indicated in Supplemental Material 3. This version significantly reduces assessment burden while still effectively covering the core dimensions of work motivation and preserving strong psychometric properties. The second-order CFA results for the short form were as follows: χ2 (229, N = 604) = 627.411, p < .001, χ2/df = 2.740, CFI = 0.950, TLI = 0.939, RMSEA = 0.054 (90% CI: 0.049–0.059), SRMR = 0.046. The standardized factor loadings of the second-order factor Extrinsic Regulation (Ext) on its subdimensions Ext_mat and Ext_soc were .964 and .867, respectively, indicating good model fit and continued support for the second-order structure of the Ext factor. These results support Hypothesis 4a. Factor loadings for the short form of the Work Motivation Inventory are presented in Table 4.
Confirmatory Factor Analysis Factor Loadings of the Short Form of the Work Motivation Inventory in Study 2
Note: Amo = Amotivation; Ext (mat/soc) = External regulation (material/social); Introj = Introjected regulation; Iden = Identified regulation; ImTK = Intrinsic motivation to know; ImAC = Intrinsic motivation to accomplish; ImEX = Intrinsic motivation to experience stimulation.
Reliability
Cronbach's alpha coefficients were calculated for each dimension of both the full version and the short form of the Work Motivation Inventory. The results are presented in Table 5. The internal consistency coefficients for the revised full version in Study 2 ranged from .721 to .911, while those for the short form ranged from .718 to .891. These results indicate that both the full version and the short form of the scale demonstrate good reliability. Table 5 also presents the intercorrelations for Study 2. Replicating the findings from Study 1, the correlations supported the proposed pattern in Hypothesis 2, where the strongest positive correlations were observed between adjacent autonomous dimensions, while Amotivation consistently showed moderate-to-strong negative correlations with all autonomous motivation subscales, providing support for Hypothesis 2.
Reliability, Descriptive Statistics, and Inter-Dimension Correlations of the Work Motivation Inventory in Study 2
Note: All correlation coefficients are significant at p < .05 level. Amo = Amotivation; Ext (mat/soc) = External regulation (material/social); Introj = Introjected regulation; Iden = Identified regulation; ImTK = Intrinsic motivation to know; ImAC = Intrinsic motivation to accomplish; ImEX = Intrinsic motivation to experience stimulation.
Criterion-Related Validity
Table 6 presents the correlations between each dimension of the Work Motivation Inventory from Study 2 and the criterion measures. The overall pattern of correlations between the work motivation dimensions and the criterion variables was consistent with theoretical expectations. Specifically, the autonomous motivation dimensions (Iden, ImTK, ImAC, and ImEX) showed significant moderate to high positive correlations with positive criteria such as Perceived Organizational Support (SPOS), Affective Commitment Scale (ACS), Satisfaction With Work Scale (SWWS), and Basic Psychological Needs (BPN), with correlation coefficients ranging from r(604) = .434 to .750 (95% CIs [.367 .783], all ps < .001). This suggests that individuals with higher levels of autonomous motivation tend to report greater feelings of organizational support, higher work satisfaction, and greater fulfillment of basic psychological needs. Furthermore, autonomous motivation dimensions were significantly negatively correlated with intention to leave (ILS), r(604) = −.596 to −.520 (95% CIs [−.645 −.459], all ps < .001), indicating that individuals with higher levels of autonomous motivation have a lower intention to leave their jobs.
Reliability, Means, Standard Deviations, and Correlations with Work Motivation Dimensions for Criterion Measures in Study 2
Note: Unless marked with ns (not significant), all correlation coefficients are significant at p < .05 level. Amo = Amotivation; Ext (mat/soc) = External regulation (material/social); Introj = Introjected regulation; Iden = Identified regulation; ImTK = Intrinsic motivation to know; ImAC = Intrinsic motivation to accomplish; ImEX = Intrinsic motivation to experience stimulation. SPOS = Survey of Perceived Organizational Support; ACS = Affective Commitment Scale; CCS = Continuance Commitment Scale; NCS = Normative Commitment Scale; SWWS = Satisfaction with Work Scale; ILS = Intent to Leave Scale; BPN (A / C / R) = Basic Psychological Needs (Autonomy / Competence / Relatedness); AMS(HS/FF) = Achievement Motives Scale (Hope of Success / Fear of Failure).
Regarding the controlled motivation dimensions, the external regulation dimensions (Ext_mat, Ext_soc) exhibited moderate positive correlations with perceived organizational support (SPOS), organizational commitment (ACS), and work satisfaction (SWWS), r(604) = .224 to .700 (95% CIs [.147 .739], all ps < .001). However, they were also significantly negatively correlated with intention to leave (ILS), r(604) = −.521 to −.315 (95% CIs [−.577 −.242], all ps < .001). This suggests that while external regulation might enhance short-term commitment or retention, its overall impact may be complex and potentially associated with other pressures or instability. Introjected regulation (Introj) was also positively correlated with positive criteria (SPOS, ACS, SWWS), r(604) = .241 to .593 (95% CIs [.164 .642], all ps < .001), although the correlation levels were generally lower compared to those of the autonomous motivation dimensions. This reflects that introjected regulation supports work motivation to some extent, but individuals driven by it may be more susceptible to the influence of external evaluations and norms.
The Hope of Success dimension of the Achievement Motivation Scale (AMS_HS) was positively correlated with the autonomous motivation dimensions, r(604) = .434 to .560 (95% CIs [.367 .613], all ps < .001), whereas the Fear of Failure dimension (AMS_FF) was negatively correlated with autonomous motivation dimensions, r(604) = −.391 to −.318 (95% CIs [−.457 −.245], all ps < .001). This result supports the link between autonomous motivation and positive achievement motivation orientations. Notably, the correlation between AMS_HS and ImAC was the highest among the autonomous dimensions r(604) = .560 (95% CI [.503 .613], p < .001), aligning with our theoretical expectations.
The correlation patterns between the dimensions of the short form of the scale and the criterion measures were largely consistent with those of the full version, providing support for Hypothesis 4b. Taken together, the Work Motivation Inventory developed in this study demonstrates good criterion-related validity.
Discussion
This study, grounded in Self-Determination Theory (SDT) and Vallerand's hierarchical model of intrinsic and extrinsic motivation, developed a novel instrument for measuring work motivation: the Work Motivation Inventory (WMI). Following two rounds of administration using different methods across diverse samples, data analysis, and revisions, the final version of the scale was established. It comprises 40 items covering seven dimensions of work motivation: amotivation, external regulation (with material and social first-order factors), introjected regulation, identified regulation, intrinsic motivation to know, intrinsic motivation to accomplish, and intrinsic motivation to experience stimulation. This structure allows for a more comprehensive assessment of individuals’ work motivation profiles. The research demonstrated that the WMI possesses good internal consistency reliability, criterion-related validity, and a stable factor structure. To meet the need for flexibility in different application contexts, a short form of the scale consisting of 24 items (3 per dimension) was also developed, effectively covering the core work motivation dimensions and maintaining sound psychometric properties while reducing the number of items and respondent burden. In summary, the Work Motivation Inventory (WMI) developed in this research serves as a valid tool for effectively measuring and evaluating individual work motivation levels in organizational and other work settings.
The Structure of Work Motivation
This research adopted Self-Determination Theory (SDT) and Vallerand's hierarchical model of intrinsic and extrinsic motivation as its theoretical foundation. Through factor analysis, a seven-factor, second-order model of work motivation was extracted and confirmed. Compared to previous research, the present study introduces the internal structure of intrinsic motivation, enabling a more comprehensive measurement of its different orientations. Furthermore, it confirmed two first-order factors within external regulation: material and social. Additionally, integrated regulation, a theoretical dimension from SDT, was excluded from the WMI's measurement model. In the following sections, we will discuss the rationale and implications of these distinctions.
The Structure of Intrinsic Motivation and the Features of Modern Organizations
In the contemporary organizational work environment, rapid technological advancements and globalization are reshaping work patterns. The rise of technologies such as artificial intelligence and big data has altered the demands placed on employees by modern enterprises: employees need not only solid technical skills and knowledge in their professional fields but also the capacity for continuous learning and innovation. The recent rapid development of Generative Artificial Intelligence (GenAI) exemplifies this trend in modern enterprises. As AI increasingly automates routine tasks, the value of human labor shifts towards solving novel problems and continuous adaptation (Korzynski et al., 2023). Consequently, the ability to learn is no longer just a preparatory phase but a core component of daily work performance. Against this backdrop, for modern enterprises to adapt, maintain innovation, and remain competitive, the requirements for employees extend beyond the technical scope. Enterprises increasingly need employees to leverage their intrinsic motivation for learning to better adapt to the needs of the organization and the job, fostering learning organizations. Employees do not cease learning upon leaving school; rather, learning becomes a lifelong process, making the relationship between work and learning more intertwined. Vallerand's hierarchical model of intrinsic and extrinsic motivation (1997) suggests that intrinsic motivation can be subdivided into three distinct orientations: to know, to accomplish, and to experience stimulation. These dimensions have been widely applied in academic contexts, and this study posits that they hold significant value in work contexts as well: employees with intrinsic motivation to know are willing to proactively explore new knowledge, driving continuous learning for employees in technology-intensive positions; employees’ intrinsic motivation to accomplish is reflected in striving for excellent performance at work, a key factor in enhancing organizational innovation and competitiveness; employees’ intrinsic motivation to experience stimulation enables sustained engagement due to the inherent enjoyment derived from the work itself, suitable for creative industries and highly autonomous roles. In the learning organizations of the new era, understanding the types of intrinsic motivation employees possess and how to foster intrinsic motivation to better align with organizational needs will become critical. Using older work motivation instruments that measure intrinsic motivation merely as a single variable may fail to meet these emerging requirements. The Work Motivation Inventory (WMI) developed in this study supports the differentiation of intrinsic motivation dimensions in work motivation measurement, allowing for a more precise revelation of employees’ motivational levels in work contexts, particularly the different orientations of intrinsic motivation, thereby aligning with the characteristics of the times and the new needs of modern enterprises.
Differentiating Material and Social Factors in External Regulation
Within SDT, external regulation refers to a type of motivation where an individual's behavior is guided by external controlling factors, such as rewards and punishments (Deci & Ryan, 1985). It is part of extrinsic motivation and lies at the most controlled end of the self-determination continuum. External regulation has typically been treated as a single dimension. However, Gagné et al. (2015) were the first to further differentiate external regulation by distinguishing between material and social external factors. Material external regulation is directly related to concrete material or financial outcomes; employees work hard to obtain economic rewards, job security, or avoid negative material consequences like unemployment. Social external regulation stems from social pressures or expectations; employees exert effort to gain respect or recognition from others, or to avoid criticism from significant others (e.g., supervisors, colleagues, family, clients). This distinction enhances conceptual precision, as the traditional concept of external regulation encompasses all externally driven behaviors, which might be overly broad. Separately measuring these factors allows researchers to understand the sources of controlled motivation in greater detail (Howard et al., 2021) and improves the instrument's ability to predict different work outcomes. For instance, excessively high material external regulation might be associated with an overemphasis on compensation, reduced job satisfaction due to insufficient rewards, or an increased risk of unethical behavior under financial pressure. Social external regulation might relate to employees’ approval-seeking behaviors towards supervisors and their anxiety or stress levels when facing criticism. Differentiating these two types helps researchers more accurately predict which type of external regulation is more strongly associated with specific work outcomes (e.g., performance, turnover intention, well-being). Although they did not differentiate intrinsic motivation, the Multidimensional Work Motivation Scale (MWMS) developed by Gagné et al. (2015) was empirically tested with large samples in up to seven languages across nine countries, demonstrating the cross-cultural consistency of the distinction within external regulation. This provides a solid foundation for the WMI developed in the present study to adopt this structure for external regulation.
Furthermore, identifying these distinct sources complements the concept of “instrumental regulation” highlighted in recent Chinese public administration research (e.g., Chen et al., 2018; Xu, 2022; Xu & Chen, 2021). While those studies emphasize the unique, quasi-autonomous nature of job security and benefits in the public sector, our model categorizes them as material external regulation based on the tangible nature of the incentives. We believe this distinction is crucial for the broader workforce, as it allows researchers to disentangle the effects of economic pragmatism (material) from the potent effects of interpersonal and cultural pressures (social), both of which are salient drivers in the Chinese context.
Removal of Integrated Regulation
Integrated regulation represents the most autonomous form of extrinsic motivation, in which a behavior is fully internalized and congruent with one's broader values and identity (Deci & Ryan, 1985; Ryan & Deci, 2000). However, its empirical distinctiveness from identified regulation and intrinsic motivation has often been difficult to establish in self-report measurement. Several SDT-based work motivation scales and validation studies have either omitted integrated regulation or noted its conceptual and empirical proximity to identified regulation and intrinsic motivation (Gagné et al., 2015; Tremblay et al., 2009; Van den Broeck et al., 2021). Similar issues have also been reported in earlier multidimensional motivation research, where adjacent autonomous forms of motivation were highly correlated or difficult to separate cleanly in factor analyses (Vallerand et al., 1992).
A similar pattern emerged in the present study. In the initial EFA, items designed to measure integrated regulation did not form a separate factor. Some integrated regulation items showed salient cross-loadings with identified regulation or intrinsic motivation factors, whereas others loaded primarily on factors that were not interpretable as a distinct integrated-regulation dimension. This pattern suggests that, in the current data, participants may have interpreted integrated regulation items as expressions of personally valuing the job or experiencing the job as meaningful, rather than as a clearly separate form of full identity integration. This may be especially relevant for Study 1, which consisted primarily of young employees in an early career training context. These participants may have been able to endorse the personal importance of work goals, but may not yet have had sufficient long-term occupational experience for work identity to become fully integrated into their broader sense of self. Therefore, to maintain structural clarity and psychometric stability, integrated regulation was not retained as a separate WMI dimension. Instead, one conceptually relevant item originally developed for integrated regulation was retained under identified regulation in the final scale.
Practical Implications
The WMI offers a versatile diagnostic tool for Human Resource (HR) professionals and managers to implement precision management beyond “one-size-fits-all” strategies. By distinguishing between specific intrinsic drives and external regulatory sources, the scale allows for the construction of individualized motivation profiles, enabling managers to tailor incentives, such as offering training for those driven by Intrinsic Motivation to Know versus public recognition for External-Social oriented employees. Beyond individual assessment, the WMI facilitates the monitoring of team climates to prevent burnout associated with Introjected Regulation and serves as a supplementary metric in recruitment and personnel selection to ensure person-job fit, such as matching candidates high in Intrinsic Motivation to Accomplish with innovation-intensive roles.
Limitations and Future Research
Although this study developed and provided initial validation evidence for the Work Motivation Inventory (WMI), several limitations should be acknowledged. To start with, all validity evidence was based on self-report data, including both WMI responses and criterion measures. Although self-report questionnaires are appropriate for assessing subjective psychological constructs such as motivation, reliance on a single data source may introduce common method bias and social desirability effects. Future research should examine the WMI using more diverse sources of evidence, such as supervisor or peer ratings, objective work performance indicators, behavioral observations, or organizational records, to further evaluate its predictive validity.
Secondly, the Study 1 sample was relatively homogeneous, consisting of young employees from a single commercial bank who were participating in a professional training program. This context represents a real HR setting in which the motivation of entry-level employees is practically meaningful; however, it may limit the generalizability of the initial item-screening results. Some candidate items referred to important work tasks, prior changes in work meaning, or achievement experiences at work, which may be interpreted differently by employees with limited work experience. Because direct tenure or banking-industry work-experience information was not collected in Study 1, we could not empirically test whether years of work experience influenced responses to these items. Several steps partly mitigated this concern: items that performed poorly in later analyses were removed during the refinement process, and Study 2 recruited a broader working sample with more substantial occupational experience (M years of work experience = 8.58, SD = 5.95, range = 1–42 years). Nevertheless, future research should further examine the WMI among employees with different levels of tenure, occupational seniority, and job responsibility, and should test measurement invariance across early career, mid career, and experienced worker groups.
Although the item pool was developed deductively from SDT, existing motivation scales, and expert review, the present study did not include qualitative interviews, focus groups, or cognitive interviews with target employees before item generation. This limits the extent to which the item pool can fully capture emic features of work motivation in Chinese organizational contexts. Several items were self-constructed to represent theoretically relevant content not adequately covered by existing instruments, but these items should be further evaluated through qualitative or mixed-method approaches. Future studies could interview employees from different industries and career stages to refine item wording, identify culturally and contextually salient forms of motivation, and examine whether additional dimensions or alternative formulations improve the WMI.
Although the WMI aims to capture a general structure of work motivation, motivation patterns and their associations with behavioral outcomes may vary across industries, organizational cultures, job characteristics, and occupational roles. For example, intrinsic motivation may be especially central in autonomous or creative jobs, whereas identified or introjected regulation may be more relevant in roles emphasizing standardized procedures or compliance. Future research should test the WMI across broader occupational groups, including employees from manufacturing, service, education, healthcare, public-sector organizations, private enterprises, managerial roles, technical positions, and frontline jobs. Such work would help clarify whether the WMI functions similarly across work contexts and whether different motivation profiles predict work outcomes in context-specific ways.
Regarding sample diversity, both samples were drawn from Chinese working populations. Cultural background may shape how employees interpret and prioritize different sources of motivation, such as social approval, material rewards, autonomy, or intrinsic enjoyment. Therefore, the structure and psychometric properties of the WMI should be tested in other national and cultural settings. Cross-cultural validation and measurement invariance testing will be important for determining whether the WMI can be used as a broader cross-cultural research tool and for identifying both universal and culture-specific aspects of work motivation.
Finally, the current study used cross-sectional designs, which limits causal interpretation and prevents assessment of temporal stability. Future research should use longitudinal designs to examine the test-retest reliability of the WMI and the dynamic changes in work motivation over time. The WMI may also be useful in intervention or quasi-experimental studies, such as pre- and post-training assessments, to evaluate programs designed to foster employees’ autonomous motivation.
Conclusions
In this study, we developed the Work Motivation Inventory (WMI) and its short form to assess differentiated work motivation profiles in Chinese working samples. Grounded in Self-Determination Theory and Vallerand's hierarchical model of intrinsic and extrinsic motivation, the WMI incorporates both the material/social distinction within external regulation and the tripartite structure of intrinsic motivation: to know, to accomplish, and to experience stimulation.
Across two studies, the WMI showed a clear and interpretable seven-factor, second-order structure, acceptable internal consistency, and meaningful associations with relevant work-related criteria, including work engagement, organizational commitment, job satisfaction, burnout, and basic psychological need satisfaction. These findings provide initial support for the WMI as a theory-guided instrument for assessing work motivation in organizational contexts. At the same time, the present evidence should be interpreted as an initial validation rather than a definitive test of the scale's generalizability. Future research should further examine the WMI across more diverse occupational groups, employees with different levels of work experience, longitudinal designs, cross-cultural samples, and qualitative or mixed-method item-development procedures.
Supplemental Material
sj-docx-1-pac-10.1177_18344909261461939 - Supplemental material for The Work Motivation Inventory: Scale Development and Initial Validation in Chinese Working Samples
Supplemental material, sj-docx-1-pac-10.1177_18344909261461939 for The Work Motivation Inventory: Scale Development and Initial Validation in Chinese Working Samples by Yiou Guo and Jian Li in Journal of Pacific Rim Psychology
Supplemental Material
sj-docx-2-pac-10.1177_18344909261461939 - Supplemental material for The Work Motivation Inventory: Scale Development and Initial Validation in Chinese Working Samples
Supplemental material, sj-docx-2-pac-10.1177_18344909261461939 for The Work Motivation Inventory: Scale Development and Initial Validation in Chinese Working Samples by Yiou Guo and Jian Li in Journal of Pacific Rim Psychology
Supplemental Material
sj-docx-3-pac-10.1177_18344909261461939 - Supplemental material for The Work Motivation Inventory: Scale Development and Initial Validation in Chinese Working Samples
Supplemental material, sj-docx-3-pac-10.1177_18344909261461939 for The Work Motivation Inventory: Scale Development and Initial Validation in Chinese Working Samples by Yiou Guo and Jian Li in Journal of Pacific Rim Psychology
Footnotes
Acknowledgements
The authors wish to express their gratitude to the following individuals for their valuable assistance during this research. We thank Qifan Yang for assisting with data collection for Study 1. We are grateful to Tian Qiu for help with organizing criterion measures and preliminary data processing for Study 1, and to Xinhai Tong for assistance with data collection for Study 1. We also appreciate Xinyi Zhu for engaging in insightful discussions that provided valuable perspectives throughout the research process.
Ethical Considerations
Ethical approval for this study was obtained from the Ethics Review Committee of the Faculty of Psychology, Beijing Normal University (IRB Number: BNU202504140109).
Consent to Participate
Written informed consent to participate was obtained from all participants. For online surveys, consent was documented by participants clicking a button to proceed after reading the information sheet.
Funding
The authors received no financial support for the research, authorship, and/or publication of this article.
Declaration of Conflicting Interests
The authors declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.
Data Availability Statement
Open Materials Statement
All materials used in this study have been uploaded as a Supplementary File for online publication only.
Supplemental Material
Supplemental material for this article is available online.
References
Supplementary Material
Please find the following supplemental material available below.
For Open Access articles published under a Creative Commons License, all supplemental material carries the same license as the article it is associated with.
For non-Open Access articles published, all supplemental material carries a non-exclusive license, and permission requests for re-use of supplemental material or any part of supplemental material shall be sent directly to the copyright owner as specified in the copyright notice associated with the article.
