Extrinsic and intrinsic rewards as strategic drivers of employee performance: The mediating role of motivation

Reward systems as a strategic HRM mechanism

In contemporary human resource management literature, the reward system is no longer treated solely as a compensation mechanism, but as a strategic practice through which the organization structures the behavior, attitudes, and outcomes of employees. The systematic review by Figueiredo et al. (2025) shows that reward systems are repeatedly linked to motivation, satisfaction, commitment, and individual performance, which places reward at the center of the contemporary HRM debate. ¹² Similarly, Li et al. (2023) argue that employees’ perception of the total reward package is not limited to monetary compensation alone but includes a broader configuration of financial and non-financial benefits, which are related to different levels of work performance. ¹³ Even the most recent sectoral evidence shows that incentive and reward systems can affect employee performance directly and indirectly, especially when mediated by job satisfaction or other organizational perceptions.^14,15 This means that the reward system should be understood as an organizational mechanism with strategic importance and not as isolated payment.

In this sense, reward systems include not only financial rewards, such as salary, bonuses and material incentives, but also non-financial elements, such as recognition, development opportunities, flexibility, and work-life balance. Li et al. (2023) show that higher performance is not necessarily associated only with compensation-dominant profiles, but also with profiles where employees report higher satisfaction with development and career opportunities or work-life balance. ¹³ In addition, Ce et al. (2025) find that monetary and non-monetary motivations affect different dimensions of job performance, ¹⁵ while Figueiredo et al. (2025) synthesize this evidence by arguing that reward systems should be treated as integrated organizational packages. ¹²

Employee motivation and work performance

Motivation is a central construct in explaining employee behavior because it relates to direction, intensity, and persistence of effort at work. Recent literature treats motivation as a mechanism that translates organizational input into commitment, involvement, and concrete work outcomes. Layek and Koodamara (2024) find that both intrinsic and extrinsic motivation are positively related to employee performance. ¹⁶ Similarly, Ce et al. (2025) show that monetary and non-monetary incentives affect financial and non-financial job performance, ¹⁵ while Elamalki et al. (2024) report that motivation is significantly related to job performance in the public sector. ¹⁷ These findings support the treatment of motivation as a close explanatory link between organizational practices and work outcomes.

Regarding performance, contemporary literature rejects the approach that reduces it only to the fulfillment of formal tasks, treating it increasingly as a multidimensional construct that includes distinct dimensions of work behavior.^18–20 Recent validations and adaptations of individual work performance instruments show that performance measurement should go beyond direct task output and include contextual and adaptive behaviors that support organizational effectiveness.^21,18 In this vein, Lousã et al. (2024) confirm the three-dimensional structure of the Individual Work Performance Questionnaire in the Portuguese version and support the measurement of performance through task performance, contextual performance and adaptive performance. ¹⁸ Additionally, Krijgsheld et al. (2024) show that adaptive performance constitutes a distinct and methodologically valuable dimension, especially in dynamic, uncertain work environments with high demands for flexibility. ¹⁹ Also, more recent studies in contemporary organizational contexts show that task, contextual and adaptive dimensions are used together to better understand employee performance under conditions of change and high complexity.^20,22 This means that employee performance should not be seen as a narrow operational outcome, but as a combination of task fulfillment, collaborative behaviors, and the ability to adapt to change. For the current study, this approach is more appropriate than limiting itself to task performance, because it allows for a more realistic and complete operationalization of performance behavior in the organization.^18,19

Theoretical grounding of the study

The theoretical basis of this study is based on the idea that the relationship between reward, motivation and performance is not linear but passes through psychological and organizational processes that structure the way employees interpret their work experience. The literature of recent years strongly supports this position. Vo et al. (2022) show that work motivation is influenced by individual needs and social conditions, suggesting that motivation is not only shaped by material gain, but also by autonomy and social connection. ²³ Similarly, Barriga Medina et al. (2024) show that motivation can play a mediating role in explaining work performance, ²⁴ while Nusraningrum et al. (2024) treat motivation as a central factor in the relationship between working conditions, engagement and performance. ²⁵ This makes the placement of motivation an explanatory mechanism in the current model theoretically sustainable.

At the same time, the literature on reward systems and incentives suggests that financial and non-financial rewards do not produce identical effects, but act in different ways and may have different consequences on motivation and performance. Figueiredo et al. (2025) show that reward systems are associated with a wide range of work outcomes and organizational behavior. ¹² Li et al. (2023) argue that different configurations of total rewards produce different performance profiles. ¹³ Alkandi et al. (2023) show that incentive and reward systems affect employee performance with job satisfaction as an intermediate mechanism, ¹⁴ while Ce et al. (2025) confirm that both monetary and non-monetary incentives are associated with different forms of job performance. ¹⁵ Theoretically, this makes it reasonable to divide the reward system into more specific dimensions and place motivation as the central link connecting them with performance. Overall, the literature from 2022 to 2025 supports three main premises for the current study: reward systems are strategic practices that influence employee behavior, motivation is a central explanatory mechanism, and performance should be treated as a multidimensional construct. This creates the conceptual and empirical basis for moving toward hypothesis development.

Theoretical framework and hypothesis development

In the human resource management literature, the reward system is treated as a strategic mechanism through which the organization orients the behavior of employees towards its objectives. It is not reduced only to pay or compensation but represents a set of policies and practices through which the organization rewards the contribution, competence, and performance of employees. From this perspective, the reward system has not only an economic function, but also a motivational one, because it communicates which behaviors, results, and standards are valued by the organization. This logic is supported by classical approaches to motivation, especially by expectancy theory, according to which the individual is more inclined to invest effort when he believes that performance will lead to desirable results and that these results will be realistically rewarded. ²⁶ Along the same lines, contemporary reward management literature argues that reward systems influence not only motivation, but also the selection, retention, and orientation of employees towards higher performance.^27,28 Empirical evidence also supports this position. Güngör (2011) argues that reward management system applications have a direct role in promoting employee performance, ²⁹ while Li et al. (2023) show that different profiles of total reward satisfaction are associated with different levels of job performance. ¹³ On this basis, the reward system should be treated as a broad strategic construct and not as an isolated payment.

A more precise analytical treatment of the reward system requires the distinction between extrinsic and intrinsic rewards. Extrinsic rewards include salaries, bonuses, fringe benefits, promotions, job security, and other forms of material or institutional benefits that originate outside the work experience itself. Their essence lies in the fact that they constitute outcomes separate from the work activity itself and are valued because of the benefit they bring to the individual. ³⁰ From the perspective of expectancy theory, extrinsic rewards increase motivation when the employee believes that the effort will produce performance, the performance will be recognized by the organization, and the recognition will be accompanied by concrete reward. ²⁶ Similarly, incentive-based perspectives suggest that when the reward is clearly linked to performance, the employee is more likely to orient his behavior towards measurable results.^31,32 The literature on merit pays and pay-for-performance also supports this argument, showing that a clear link between performance and material reward can increase motivation, especially when the system is perceived as fair, transparent, and trustworthy. ³³ Güngör (2011) also treats financial rewards as a key source of extrinsic motivation. ²⁹ This means that extrinsic rewards do not only have a compensatory function but can act as external incentives that push the employee towards higher effort. On this basis, it is formulated:

H₁

Extrinsic rewards have a positive impact on employee motivation

Intrinsic rewards, on the other hand, stem from the work experience itself and from the way the employee experiences it. They include recognition, appreciation, autonomy, involvement in decision-making, constructive feedback, opportunities for learning and development, and a sense that work has meaning and importance. These forms of reward are particularly important within the framework of self-determination theory, which distinguishes between intrinsic and extrinsic motivation and argues that individuals are more deeply engaged when their needs for autonomy, competence, and social connection are met.^30,34 In the same vein, the job characteristics model argues that the characteristics of the job itself, such as autonomy, feedback, and meaningfulness, directly affect the employee’s intrinsic motivation. ³⁵ Turner (2017) argues that the performance culture in an organization is not built only on external incentives, but also on intrinsic motivation, ³⁶ while Kuvaas et al. (2017) show that intrinsic motivation is more consistently associated with positive employee outcomes than extrinsic motivation. ³⁷ This implies that non-financial forms of reward are not secondary elements, but central factors in building intrinsic motivation. For this reason, it is formulated:

H₂

Intrinsic rewards have a positive impact on employee motivation.

Literature supports not only an indirect link between extrinsic rewards and performance through motivation, but also a direct effect. From the perspective of reinforcement theory, behavior is influenced by its consequences, and when performance is explicitly rewarded, the likelihood of its repetition increases. ³⁸ Similarly, expectancy theory and incentive-based perspectives argue that when employees perceive rewards as linked to achievement, they are more likely to direct their behavior towards valued outcomes.^26,32 Empirically, Güngör (2011) shows that financial rewards appear as an important component of reward management system applications in relation to employee performance, ²⁹ while Nyberg et al. (2016) show that merit pay and bonus pay are positively related to future employee performance. ³³ Although these effects depend on the fairness, clarity, and structure of the system, literature provides sufficient basis to expect a positive effect of extrinsic rewards on performance. Thus, it is formulated:

H₃

Extrinsic rewards have a positive impact on employee performance.

At the same time, intrinsic rewards also have a strong theoretical basis for being directly linked to performance. The job characteristics model argues that when work is characterized by autonomy, meaning, and feedback, employees experience higher intrinsic motivation and perform better. ³⁵ Similarly, self-determination theory emphasizes that meeting the needs for competence, autonomy, and social connection supports deeper engagement and more sustained performance.^30,39 Empirically, Li et al. (2023) show that different profiles of total reward satisfaction produce different levels of task performance and contextual performance, and that higher results are not necessarily related only to compensation, but also to development opportunities and work-life balance. ¹³ This suggests that non-financial dimensions of the reward system can directly influence performance, because they strengthen the employee’s sense of appreciation, competence, and involvement. Therefore, it is formulated:

H₄

Intrinsic rewards have a positive impact on employee performance.

Motivation occupies a central place in this model, because it represents the psychological mechanism that transforms incentive into effort and effort into results. In the classical theory of Lawler and Porter (1967), motivation, satisfaction and performance are treated as interrelated, but not identical factors, ⁴⁰ while Vroom (1964) explains motivation as the result of the individual’s assessment of the relationship between effort, performance and reward. ²⁶ In this sense, motivation is not simply a passive psychological state, but a force that guides the intensity, direction and sustainability of employee behavior. Empirical literature strongly supports this connection. Güngör (2011) places motivation at the center of his model, arguing that reward systems are used precisely to increase motivation and, through it, performance. ²⁹ Layek and Koodamara (2024) find strong positive relationships between intrinsic motivation, extrinsic motivation and employee performance, ¹⁶ while Deci, Olafsen, and Ryan (2017) argue that more autonomous forms of motivation are more consistently associated with higher performance and better functioning in the workplace. ³⁹ This makes motivation a central determinant of performance. Therefore, it is formulated:

H₅

Motivation has a positive impact on employee performance.

The strongest theoretical point of the model is that motivation mediates the relationship between rewards and performance. In Güngör’s (2011) model, the reward management system influences employee performance not only directly, but also through motivation. ²⁹ For extrinsic rewards, this logic is particularly clear, because salary, bonuses and performance-based incentives create external expectations that increase motivation to achieve. From the perspective of expectancy theory, the employee increases effort when he believes that high performance will lead to valuable rewards (Vroom, 1964), while equity theory suggests that motivation will be higher when the ratio between inputs and outputs is perceived as fair compared to others (Adams, 1965). This implies that the impact of extrinsic rewards on performance is not expected to be only linear, but to pass, at least partially, through motivation as an explanatory mechanism. On this basis, it is formulated:

H₆

Motivation mediates the relationship between extrinsic rewards and employee performance.

The same logic applies to intrinsic rewards. Non-financial rewards, such as recognition, autonomy, feedback, and development opportunities, do not affect performance simply because they are perceived as rewards, but because they change the way an employee experiences their work and their role in the organization. In the context of self-determination theory, these elements increase intrinsic motivation by fulfilling the individual’s basic psychological needs.^34,30 Similarly, Hackman and Oldham (1976) argue that job characteristics create conditions for intrinsic motivation, which is then reflected in higher performance. ³⁵ This logic is also supported by empirical literature, Güngör (2011) specifically links non-financial rewards to intrinsic motivation, ²⁹ while Kuvaas et al. (2017) show that intrinsic motivation is more strongly associated with positive employee outcomes. ³⁷ Consequently, intrinsic rewards are expected to initially increase the employee’s internal motivation and then, through this mechanism, be reflected in higher performance. This is how it is formulated:

H₇

Motivation mediates the relationship between intrinsic rewards and employee performance.

Regarding the dependent variable, the literature defines individual work performance as a multidimensional construct. Koopmans et al. (2011) identify four central dimensions: task performance, contextual performance, adaptive performance and counterproductive work behavior. ⁴¹ In the current study, the use of the first three positive dimensions is methodologically reasonable, because they cover the fulfillment of basic tasks, cooperative behaviors and the ability to adapt to changes. Furthermore, Koopmans et al. (2014) identify the most relevant indicators for each dimension and show that performance evaluation should not be limited to formal work output alone. ⁴² For this reason, employee performance in this study is conceptualized through task performance, contextual performance and adaptive performance, as three positive dimensions that more fully reflect performance behavior in the organization.

Based on the theoretical arguments and empirical evidence discussed above, the following conceptual framework was developed to examine the relationships between extrinsic rewards, intrinsic rewards, motivation, and employee performance. The model proposes that both reward dimensions influence employee performance directly and indirectly through the mediating role of motivation. Figure 1 presents the hypothesized structural relationships examined in this study.OPEN IN VIEWER

Respondents and sampling procedure

The target population of this study consisted of employees of private enterprises in Kosovo, while the unit of analysis was the individual employee. The study focused on the private sector because reward, motivation and performance practices are directly related to the way enterprises organize and manage their human resources.

Respondents were selected through purposive non-probability sampling. This approach was used because the study required participants employed in private enterprises who were able to evaluate reward practices, motivation, and performance within their organizational context. To ensure broader variation, the sample included 30 private enterprises from different sectors, including services, trade, manufacturing, finance, administration, and information technology. In terms of territorial distribution, the enterprises were selected from several economic centers of Kosovo.

The final sample included 300 respondents. To reach this number of valid responses, the initial distribution of 375 questionnaires was planned, considering an expected non-response rate of 20%. The required number of questionnaires was determined using the formula:

N N R = n 1 − r = 300 1 − 0.20 = 375

This indicated that 375 questionnaires needed to be distributed, corresponding to approximately 12–13 questionnaires per enterprise, with the aim of obtaining, on average, about 10 valid responses from each enterprise. The sample size was considered appropriate for testing the measurement and structural model, which includes four constructs: extrinsic rewards, intrinsic rewards, motivation, and employee performance. The dependent construct, employee performance, receives three direct paths from extrinsic rewards, intrinsic rewards and motivation.

In demographic terms, 56.0% of the respondents were female and 44.0% were male. The most represented age group was 25–34 years, accounting for 32.0% of the sample, while respondents aged 35–44 years accounted for 26.0%. Regarding education, 42.0% had completed bachelor’s studies and 32.0% had completed master’s studies. In terms of work experience, 27.0% had more than 10 years of experience and 26.0% had 4–7 years of experience.

Measurement scales

The research instrument was developed in accordance with the conceptual model of the study, which examines the relationship between reward systems, motivation and employee performance. The questionnaire was organized into two main sections. The first section included demographic questions related to gender, age, education level, work experience and monthly net salary. The second section included the items measuring the four study constructs: extrinsic rewards, intrinsic rewards, motivation and employee performance.

The construction of the instrument was guided by the reward-motivation-performance logic proposed by Güngör (2011). In the present study, the reward system was operationalized through two dimensions: extrinsic rewards and intrinsic rewards. Extrinsic rewards were measured with 10 items reflecting financial and material aspects of rewards, including fair compensation, bonuses, promotion opportunities, competitive salary, job security, additional benefits, transparency of reward policies, overtime compensation and additional compensation for exceptional engagement. Intrinsic rewards were measured with 10 items reflecting non-financial rewards, including verbal recognition, respect from management, skills development, meaningful work, participation in decision-making, autonomy, constructive feedback, responsibility, personal and professional fulfillment and symbolic recognition. Regarding motivation was treated as a mediating construct and was measured with 5 items. These items captured employees’ willingness to give maximum effort, the motivational role of achievement-based rewards, perceived organizational investment in employee development, clear communication of reward-related goals and the perceived link between individual effort and received rewards.

Employee performance was measured with a 12-item self-report scale adapted from the multidimensional logic of individual work performance proposed by Koopmans et al. (2014). In this study, employee performance covered three dimensions: task performance, contextual performance and adaptive performance. Task performance was measured through items related to completing work within deadlines, working accurately, achieving objectives and organizing tasks efficiently. Contextual performance was measured through items related to helping colleagues, taking initiative, maintaining professional and cooperative behavior and accepting responsibilities beyond minimum job requirements. Adaptive performance was captured through items related to adapting to changes, dealing effectively with unexpected situations, learning new work methods and maintaining stable performance under pressure.

All study construct items were formulated as statements and measured using a five-point Likert scale ranging from 1 = strongly disagree to 5 = strongly agree. Higher scores indicated stronger agreement with the respective statements and a higher perceived level of the measured construct. The items were adapted and contextualized for employees of private enterprises in Kosovo while preserving the theoretical meaning of the original constructs.

Pilot testing and preliminary reliability assessment

A pilot test was conducted before the main survey administration because the questionnaire was adapted and contextualized for the purpose of this study rather than used as a fully standardized instrument. The aim of the pilot phase was to examine whether the items were understandable, whether the questionnaire structure was logical and whether the measurement scale was suitable for employees of private enterprises in Kosovo. The pilot study included 30 employees from private enterprises with characteristics similar to the target respondents. These participants were not included in the final sample.

The pilot phase focused on three aspects of the instrument. First, face validity was assessed to determine whether the items appeared clear, relevant and appropriate for the intended respondents. Second, the structure of the questionnaire was reviewed to assess the logical order of the sections and the flow from demographic questions to the study constructs. Third, preliminary internal consistency was examined using Cronbach’s alpha for each construct: extrinsic rewards, intrinsic rewards, motivation, and employee performance.

The pilot results indicated that the questionnaire was generally understandable and suitable for the study context. Based on the feedback received, minor linguistic and technical adjustments were made to improve clarity and readability. Since the pilot sample was limited, Cronbach’s alpha was used only as an initial reliability check, while the main purpose of the pilot test was to refine the questionnaire before its administration to the final sample.

Data collection process

Primary data was collected through a structured questionnaire administered to employees of private enterprises in Kosovo. The data collection was conducted across 30 private enterprises operating in different sectors, including services, trade, manufacturing, finance, administration, and information technology. This approach was used to obtain responses from employees working in diverse organizational settings within the private sector.

Before completing the questionnaire, respondents were informed about the purpose of the study and the academic nature of the research. Participation was voluntary, and respondents were assured that their answers would be treated anonymously and confidentially. No financial or material incentives were offered for participation.

A total of 375 questionnaires were distributed, based on the planned final sample size and the expected non-response rate. After the data collection process was completed, 300 valid questionnaires were retained. The completed questionnaires were reviewed for completeness, coded and organized for subsequent statistical analysis.

Data screening and preparation

Before conducting the main statistical analyses, the dataset was screened for missing values, outliers and coding errors. Frequency distributions were examined for demographic variables, while descriptive statistics were calculated for the main constructs. The values of mean, standard deviation, median, minimum, maximum, skewness and kurtosis were used to assess the distribution of the data.

Normality was assessed using skewness and kurtosis values. The results showed that skewness ranged from −0.15 to 0.19 and kurtosis ranged from −0.66 to −0.38, indicating that the data did not show serious deviations from normality. This supported the use of further multivariate analyses, including Exploratory Factor Analysis, Confirmatory Factor Analysis and Structural Equation Modelling.

Statistical analysis

The data were analyzed using SPSS and RStudio. Before conducting the main analyses, the dataset was checked for missing values, coding errors and inconsistent response patterns. Only complete and usable questionnaires were included in the final analysis. Descriptive statistics were used to present the demographic profile of the respondents and to summarize the main study constructs. Frequencies and percentages were reported for categorical variables such as gender, age, education, work experience, and salary level, while means, standard deviations, skewness and kurtosis were calculated for the latent constructs.

The measurement instrument was first examined through Exploratory Factor Analysis. Before factor extraction, the adequacy of the data was assessed using the Kaiser–Meyer–Olkin measure and Bartlett’s test of sphericity. The KMO value was 0.883, indicating very good sampling adequacy, and Bartlett’s test was statistically significant, χ²(496) = 2341.697, p < .001. These results confirmed that the data were suitable for factor analysis. Principal Axis Factoring with Promax rotation was applied, as the constructs were expected to be conceptually related. The analysis produced a four-factor solution corresponding to extrinsic rewards, intrinsic rewards, motivation, and employee performance. In total, 34 items were retained, all exceeding the minimum factor loading threshold of 0.40.

After the exploratory stage, Confirmatory Factor Analysis was used to assess whether the four-factor structure was supported by the data. The standardized factor loadings were positive and statistically significant at p < .001, indicating that the observed items were meaningfully related to their respective latent constructs. Although some loadings were moderate, they were retained because they were theoretically consistent with the constructs and acceptable for the purposes of the study. Model fit was evaluated using commonly reported fit indices, including χ²/df, CFI, TLI, RMSEA, and SRMR.

Reliability and validity were then examined before testing the structural relationships. Internal consistency was assessed using Cronbach’s Alpha and Composite Reliability. Cronbach’s Alpha values ranged from 0.737 to 0.813, while Composite Reliability values ranged from 0.740 to 0.815, exceeding the recommended threshold of 0.70. Convergent validity was examined through Average Variance Extracted. Although the AVE values were below the preferred threshold of 0.50, the constructs were retained because the reliability values were acceptable and all CFA loadings were statistically significant. Discriminant validity was assessed through inter-construct correlations, which showed moderate associations and supported the empirical distinction between the constructs.

The hypothesized relationships were tested using Structural Equation Modelling. The structural model examined the effects of extrinsic and intrinsic rewards on motivation, the effect of motivation on employee performance and the direct effects of both reward dimensions on employee performance. Multicollinearity was checked through Variance Inflation Factor values. Since the VIF values ranged from 1.142 to 1.258, they were well below the threshold of 5, indicating that multicollinearity did not represent a methodological concern.

Mediation analysis was also conducted to assess whether motivation explained the relationship between reward dimensions and employee performance. The indirect effects were estimated through the product of the relevant structural paths and evaluated using standard errors, z-values, p-values and 95% confidence intervals. This allowed the analysis to capture both the direct effects of reward dimensions and the indirect pathway through motivation.

Descriptive statistics and normality assessment

Table 1 presents the descriptive statistics for the main constructs of the study. The mean values are relatively close to each other and range from 3.15 to 3.27, which suggests that respondents generally reported moderate levels of perceived rewards, motivation and employee performance. Among the constructs, employee performance had the highest mean score (M = 3.27, SD = 0.77), indicating that respondents evaluated their work performance slightly more positively compared to the other dimensions. Intrinsic rewards followed with a mean of 3.22 (SD = 0.69), while extrinsic rewards and motivation had slightly lower mean values, 3.16 and 3.15, respectively.

OPEN IN VIEWER

Table 1.

Descriptive statistics and normality assessment.

Construct	N	M	SD	Median	Min	Max	Range	Skewness	Kurtosis	SE
Extrinsic rewards	300	3.16	0.75	3.2	1.3	5	3.7	−0.14	−0.56	0.04
Intrinsic rewards	300	3.22	0.69	3.2	1.4	4.8	3.4	−0.15	−0.64	0.04
Motivation	300	3.15	0.84	3	1.4	5	3.6	0.19	−0.66	0.05
Employee performance	300	3.27	0.77	3.22	1.33	5	3.67	0.05	−0.38	0.04

The standard deviations show a moderate level of variation in the responses. Motivation had the highest standard deviation (SD = 0.84), which indicates that respondents differed somewhat more in how they perceived their level of motivation. In contrast, intrinsic rewards had the lowest standard deviation (SD = 0.69), suggesting more consistent responses for this construct. The skewness and kurtosis values were within acceptable ranges, with skewness ranging from −0.15 to 0.19 and kurtosis from −0.66 to −0.38. This indicates that the data did not show serious deviations from normality and were suitable for further analyses such as EFA, CFA, and SEM.

Exploratory factor analysis

Prior to conducting CFA, Exploratory Factor Analysis was performed to examine the underlying structure of the measurement instrument. The Kaiser–Meyer–Olkin value was 0.883, indicating very good sampling adequacy. Bartlett’s test of sphericity was statistically significant, χ²(496) = 2341.697, p < .001, confirming that the correlation matrix was appropriate for factor extraction. Principal Axis Factoring with Promax rotation was applied, as the constructs were theoretically expected to be correlated. The analysis extracted four factors, consistent with the conceptual model of the study, and retained 34 items. The four-factor solution explained 31.96% of the total variance, providing an acceptable basis for proceeding with CFA and further model validation (Table 2).

OPEN IN VIEWER

Table 2.

KMO, Bartlett’s test, and explained variance.

Indicator	Value
Kaiser–Meyer–Olkin Measure of Sampling Adequacy	0.883
Bartlett’s test of sphericity, χ2	2341.697
Df	496
p-value	<.001
Extraction method	Principal Axis Factoring
Rotation method	Promax with Kaiser Normalization
Number of extracted factors	4
Number of retained items	34
Total variance explained	31.96%

The Pattern Matrix in Table 3 indicated a four-factor solution consistent with the proposed measurement structure. All retained items exceeded the minimum loading threshold of 0.40, with factor loadings ranging from 0.413 to 0.648. Employee performance items loaded between 0.450 and 0.628, extrinsic rewards between 0.470 and 0.606, intrinsic rewards between 0.413 and 0.588, and motivation between 0.503 and 0.648. These values suggest that the retained items contributed adequately to their respective latent constructs. Although the factor loadings were generally moderate rather than high, the pattern of loadings was theoretically coherent and empirically interpretable. The results therefore support the empirical separation of employee performance, extrinsic rewards, intrinsic rewards, and motivation as four related but distinct constructs. Based on these findings, the retained 34-item, four-factor solution was considered appropriate for subsequent validation through CFA.

OPEN IN VIEWER

Table 3.

Exploratory factor analysis pattern matrix.

Construct	Item	Factor loading
Employee performance	I complete my work within the established deadlines.	0.628
	My work is accurate and free from unnecessary errors.	0.604
	I achieve established work objectives and standards.	0.600
	I organize tasks in a way that increases my efficiency at work.	0.595
	I provide help to colleagues when they need it without waiting for a formal request.	0.554
	I take initiative to improve teamwork processes.	0.532
	I maintain professional and cooperative behavior even in difficult situations.	0.479
	I am willing to take responsibility beyond the minimum duties of my position.	0.464
	I adapt quickly to changes in work processes, tools, or job requirements.	0.455
Extrinsic rewards	I receive fair compensation for the work I perform in the organization.	0.606
	Bonuses and monetary rewards are distributed fairly within the organization.	0.589
	Promotion at work is based on measurable performance rather than personal preferences.	0.572
	Salaries in this organization are competitive compared to the sector.	0.563
	I feel secure in my position due to clear employment policies.	0.551
	The organization offers additional benefits such as insurance, paid leave, or flexible working hours.	0.555
	Financial reward policies are transparent and understandable to everyone.	0.555
	Additional payments for overtime or work under special conditions are provided regularly.	0.55
	I receive additional compensation for exceptional commitment or specific achievements.	0.47
Intrinsic rewards	I receive verbal recognition and appreciation for good results at work.	0.588
	I feel respected and valued by management.	0.527
	I have opportunities to develop my skills through training and work experience.	0.526
	I am involved in projects where I feel that my work is really important for the organization.	0.515
	I have the opportunity to propose ideas and participate in team decision-making.	0.487
	I have freedom to organize my tasks and daily priorities.	0.464
	The feedback I receive from supervisors is useful and oriented toward improvement.	0.463
	The responsibilities assigned to me make me feel engaged and motivated.	0.452
	The work I perform gives me a sense of personal and professional fulfillment.	0.449
	I receive public or symbolic recognition for the good work I do.	0.413
Motivation	I feel motivated to give my maximum effort at work.	0.648
	My motivation increases when rewards are linked to achievements.	0.613
	I feel that the organization invests in my development.	0.548
	Clear communication of reward objectives motivates me.	0.503

The correlation matrix between factors showed positive and moderate correlations between the four extracted factors. The correlation values ranged from 0.345 to 0.495, indicating that the factors are interrelated, but not at a level that suggests high overlap between them. The highest correlation was evidenced between Factor 1 and Factor 2 (r = 0.495), while the lowest correlation was evidenced between Factor 2 and Factor 4 (r = 0.345). These results are consistent with the use of Promax rotation, as the factors were expected to be interrelated. At the same time, the lack of very high correlations indicates that the factors retain sufficient empirical distinctness. Therefore, the results of the correlation matrix support the four-factor structure identified in the EFA and provide a suitable basis for proceeding with confirmatory factor analysis (Table 4).

OPEN IN VIEWER

Table 4.

Factor correlation matrix.

Factor	1	2	3	4
1	1.000	​	​	​
2	.495	1.000	​	​
3	.467	.398	1.000	.418
4	.479	.345	.418	1.000

Confirmatory factor analysis and measurement model evaluation

Table 5 reports on the standardized factor loadings of the CFA measurement model. All indicators loaded positively on their respective constructs and were statistically significant at p < .001, providing initial support for the specification of the four latent variables. The standardized loading ranged from 0.423 to 0.697, which indicates that the items were meaningfully associated with their intended constructs, although the strength of these associations varied across dimensions.

OPEN IN VIEWER

Table 5.

Standardized factor loadings.

Construct	Item	Loading	SE	z	p-value
Extrinsic rewards	ER1	0.552	0.044	12.41	<.001
	ER2	0.607	0.044	13.766	<.001
	ER3	0.596	0.05	11.987	<.001
	ER4	0.55	0.043	12.756	<.001
	ER5	0.423	0.05	8.401	<.001
	ER6	0.555	0.048	11.534	<.001
	ER7	0.493	0.052	9.423	<.001
	ER8	0.54	0.046	11.655	<.001
	ER9	0.612	0.043	14.224	<.001
	ER10	0.579	0.048	12.038	<.001
Intrinsic rewards	IR1	0.496	0.051	9.79	<.001
	IR2	0.557	0.052	10.79	<.001
	IR3	0.488	0.052	9.298	<.001
	IR4	0.511	0.054	9.461	<.001
	IR5	0.488	0.055	8.902	<.001
	IR6	0.484	0.051	9.555	<.001
	IR7	0.439	0.056	7.797	<.001
	IR8	0.483	0.051	9.393	<.001
	IR9	0.498	0.055	9.032	<.001
	IR10	0.531	0.049	10.745	<.001
Motivation	MOT1	0.592	0.046	12.791	<.001
	MOT2	0.563	0.052	10.895	<.001
	MOT3	0.697	0.043	16.306	<.001
	MOT4	0.543	0.049	11.014	<.001
	MOT5	0.606	0.047	12.937	<.001
Employee performance	PERF1	0.582	0.048	12.112	<.001
	PERF2	0.633	0.038	16.605	<.001
	PERF3	0.601	0.044	13.78	<.001
	PERF4	0.515	0.044	11.675	<.001
	PERF5	0.549	0.044	12.369	<.001
	PERF6	0.598	0.045	13.263	<.001
	PERF7	0.566	0.047	12.163	<.001
	PERF8	0.566	0.044	12.781	<.001
	PERF9	0.499	0.051	9.849	<.001

For extrinsic rewards, the loadings ranged from 0.423 to 0.612, with most indicators clustering around the moderate range. This suggests that the construct captures several related but not identical aspects of external reward practices, including compensation, bonuses, promotion, benefits, and reward transparency. A similar pattern was observed for intrinsic rewards, where the loading ranged from 0.439 to 0.557. These values indicate that non-financial reward experiences such as recognition, autonomy, feedback and development opportunities contributed to the construct, but with relatively moderate individual strength.

The measurement of motivation showed comparatively stronger loadings, ranging from 0.543 to 0.697. This indicates a clearer association between the motivation indicators and the underlying latent construct, particularly for the item related to organizational investment in employee development. Employee performance also showed a relatively stable pattern of loadings, ranging from 0.499 to 0.633, suggesting that the performance indicators consistently reflected the behavioral outcomes included in the model. The CFA results indicate an acceptable measurement structure. Although several loadings did not reach the preferred level of 0.70, all indicators were statistically significant and theoretically consistent with their constructs. On this basis, the measurement model was retained for further assessment of reliability, convergent validity, discriminant validity, and subsequent structural relationships.

Table 6 presents the reliability and convergent validity indicators for the measurement model. Cronbach’s alpha values ranged from 0.737 to 0.813, and composite reliability values ranged from 0.740 to 0.815, exceeding the recommended threshold of 0.70 across all constructs. These findings indicate acceptable internal consistency and suggest that the indicators within each construct measure a coherent latent dimension. The strongest reliability was observed for extrinsic rewards and employee performance, while motivation and intrinsic rewards also demonstrated satisfactory reliability.

OPEN IN VIEWER

Table 6.

Reliability and convergent validity.

Construct	Cronbach’s Alpha	CR	AVE	√AVE
Extrinsic rewards	0.813	0.815	0.307	0.554
Intrinsic rewards	0.767	0.767	0.248	0.498
Motivation	0.737	0.74	0.364	0.603
Employee performance	0.81	0.812	0.324	0.569

The AVE values ranged from 0.248 to 0.364, indicating that the average variance extracted by the constructs was below the conventional benchmark of 0.50. However, this result should be interpreted together with the reliability of coefficients and CFA estimates. Since all composite reliability values were above 0.70 and all standardized factor loadings were statistically significant at p < .001, the constructs retained sufficient measurement consistency for further analysis. Therefore, although the AVE values suggest modest convergent strength, the overall reliability and CFA evidence support the retention of the measurement model for testing the structural relationships.

Table 7 reports the model fit indices for both the CFA and SEM models. The results indicate a strong fit between the proposed model and the empirical data. The chi-square value was non-significant, χ² (521) = 556.618, p = 0.136, while the relative chi-square was very low, χ²/df = 1.068, indicating that the discrepancy between the observed and model-implied covariance matrices was minimal.

OPEN IN VIEWER

Table 7.

CFA and SEM model fit indices.

Fit index	CFA model	SEM model
χ 2	556.618	556.618
Scaled χ2	553.321	553.321
Df	521	521
p-value	0.136	0.136
χ 2 /df	1.068	1.068
CFI	0.983	0.983
TLI	0.982	0.982
RMSEA	0.015	0.015
RMSEA 90% CI	[0.000, 0.025]	[0.000, 0.025]
SRMR	0.043	0.043

The incremental fit indices were also highly satisfactory, with CFI = 0.983 and TLI = 0.982 for both models. These values are well above the commonly accepted threshold of 0.90 and also exceed the stricter criterion of 0.95, indicating very good comparative model fit. The absolute fit indices further support this conclusion, with RMSEA = 0.015, 90% CI [0.000, 0.025] and SRMR = 0.043. The very low RMSEA value, together with its narrow confidence interval, suggests a close approximation of the model to the data, while the SRMR value indicates low standardized residuals. The identical fit indices for the CFA and SEM models show that the transition from the measurement model to the structural model did not reduce model fit. These results provide strong empirical support for the adequacy of the proposed model and justify the interpretation of the structural paths among rewards, motivation, and employee performance.

Table 8 presents the discriminant validity assessment based on inter-construct correlations, the square root of AVE and HTMT values. The correlations among the constructs ranged from 0.432 to 0.630, indicating positive and moderate associations between the dimensions of the model. The highest correlation was observed between motivation and employee performance (r = 0.630), which is consistent with the conceptual expectation that more motivated employees tend to report higher levels of performance.

OPEN IN VIEWER

Table 8.

Discriminant validity assessment using correlations, square root of AVE, and HTMT.

Construct pair	Correlation	√AVE 1	√AVE 2	HTMT
ER – IR	0.447	0.554	0.498	0.445
ER – MOT	0.432	0.554	0.603	0.449
ER – PERF	0.54	0.554	0.569	0.537
IR – MOT	0.518	0.498	0.603	0.518
IR – PERF	0.508	0.498	0.569	0.51
MOT – PERF	0.63	0.603	0.569	0.637

The Fornell-Larcker results show that the square root of AVE was generally close to the corresponding inter-construct correlations, with some pairs presenting marginal overlap. However, the HTMT results provide stronger evidence for discriminant validity, as all HTMT values ranged from 0.445 to 0.637, remaining well below the conservative threshold of 0.85. This indicates that the constructs are empirically related, but sufficiently distinct from one another. The discriminant validity results support the adequacy of the measurement model. Although the constructs share meaningful conceptual relationships, particularly motivation and employee performance, the HTMT values confirm that the latent variables do not represent overlapping dimensions. Therefore, the measurement model was considered suitable for the subsequent structural model analysis.

Structural model assessment and direct effects

Figure 1 presents the standardized estimates of the structural equation model. The measurement component shows positive factor loadings for all indicators on their respective latent constructs, consistent with the CFA results reported above. The structural component presents the five direct paths included in the model: extrinsic rewards to motivation, intrinsic rewards to motivation, motivation to employee performance, extrinsic rewards to employee performance and intrinsic rewards to employee performance.

The standardized coefficients show positive paths from extrinsic rewards to motivation (β = 0.250) and employee performance (β = 0.285), as well as from intrinsic rewards to motivation (β = 0.406) and employee performance (β = 0.162). Motivation also showed a positive path to employee performance (β = 0.423). Among the direct structural paths, the strongest standardized coefficient was observed for the path from motivation to employee performance, followed by the path from intrinsic rewards to motivation. Figure 1 therefore provides a visual summary of the direct structural relationships, while the statistical significance of these paths is reported in Table 9 (Figure 2).

OPEN IN VIEWER

Table 9.

Direct effects and structural path coefficients.

Hypothesis	Direct path	Estimate	Β	SE	z	p-value	95% CI	Decision
H1	ER → MOT	0.262	0.25	0.091	2.893	0.004	[0.085, 0.440]	Supported
H2	IR → MOT	0.473	0.406	0.11	4.314	<.001	[0.258, 0.687]	Supported
H3	ER → PERF	0.293	0.285	0.094	3.125	0.002	[0.109, 0.478]	Supported
H4	IR → PERF	0.185	0.162	0.105	1.772	0.076	[-0.020, 0.390]	Not supported
H5	MOT → PERF	0.416	0.423	0.098	4.234	<.001	[0.223, 0.608]	Supported

OPEN IN VIEWER

Figure 2.

Standardized structural equation model results.

Based on the standardized coefficients presented in Figure 2, the structural relationships can be expressed as follows:

M O T = 0.250 E R + 0.406 I R + ε 1

P E R F = 0.285 E R + 0.162 I R + 0.423 M O T + ε 2

Table 9 reports the direct effects among the latent constructs included in the structural model. The results show that extrinsic rewards had a positive and statistically significant effect on motivation (β = 0.250, z = 2.893, p = 0.004, 95% CI [0.085, 0.440]), supporting H1. Intrinsic rewards also had a positive and statistically significant effect on motivation (β = 0.406, z = 4.314, p < .001, 95% CI [0.258, 0.687]), supporting H2. The higher standardized coefficient for intrinsic rewards indicates a stronger direct association with motivation compared with extrinsic rewards.

Regarding employee performance, extrinsic rewards showed a positive and statistically significant direct effect (β = 0.285, z = 3.125, p = 0.002, 95% CI [0.109, 0.478]), supporting H3. In contrast, the direct effect of intrinsic rewards on employee performance was positive but not statistically significant (β = 0.162, z = 1.772, p = 0.076, 95% CI [−0.020, 0.390]); therefore, H4 was not supported. Motivation had a positive and statistically significant effect on employee performance (β = 0.423, z = 4.234, p < .001, 95% CI [0.223, 0.608]), supporting H5. The direct effects indicate that both reward dimensions had significant positive effects on motivation. Among the direct paths to employee performance, motivation showed the strongest standardized effect, followed by extrinsic rewards, while the direct effect of intrinsic rewards did not reach statistical significance. This pattern indicates that the contribution of intrinsic rewards to employee performance is further assessed through the mediation analysis reported in the following section.

Explanatory power and multicollinearity diagnostics

Table 10 reports on the explanatory power of the structural model and the multicollinearity diagnostics. The results show that extrinsic rewards and intrinsic rewards explained 31.8% of the variance in motivation (R² = 0.318). In addition, extrinsic rewards, intrinsic rewards and motivation explained 50.3% of the variance in employee performance (R² = 0.503). These findings indicate that the model explains a moderate proportion of variance in the endogenous constructs, with stronger explanatory power for employee performance than for motivation.

OPEN IN VIEWER

Table 10.

Explanatory power and multicollinearity diagnostics.

Endogenous construct	Predictor	R2	VIF
Motivation	ER	0.318	1.142
Motivation	IR	0.318	1.142
Employee performance	ER	0.503	1.214
Employee performance	IR	0.503	1.258
Employee performance	MOT	0.503	1.253

The VIF values ranged from 1.142 to 1.258, remaining well below the recommended threshold of 5. This indicates that multicollinearity was not a concern among the predictor constructs. Therefore, the structural path coefficients can be interpreted without evidence of distortion caused by collinearity.

Mediation analysis

After assessing the explanatory power and collinearity diagnostics, mediation analysis was conducted to examine whether motivation mediated the relationships between rewards and employee performance. The indirect effects were calculated using the standardized path coefficients as the product of the path from the independent variable to motivation and the path from motivation to employee performance. Following the assessment of the direct structural relationships, explanatory power and multicollinearity diagnostics, mediation analysis was conducted to examine whether motivation mediated the relationships between rewards and employee performance. The indirect effects were calculated using the standardized path coefficients. Specifically, the indirect effect was estimated as the product of the path from the independent variable to motivation and the path from motivation to employee performance:

I n d i r e c t E f f e c t = a × b

I n d i r e c t E f f e c t E R = 0.250 × 0.423 = 0.106

I n d i r e c t E f f e c t I R = 0.406 × 0.423 = 0.172

The total effects were calculated as the sum of the direct and indirect effects:

T o t a l E f f e c t = D i r e c t E f f e c t + I n d i r e c t E f f e c t

T o t a l E f f e c t E R = 0.285 + 0.106 = 0.391

T o t a l E f f e c t I R = 0.162 + 0.172 = 0.334

The mediation results, including the indirect and total effects, are presented in Table 11. The indirect effect of extrinsic rewards on employee performance through motivation was positive and statistically significant (β = 0.106, z = 2.531, p = 0.011, 95% CI [0.025, 0.193]), supporting H6. Since the direct effect of extrinsic rewards on employee performance was also statistically significant, this result indicates partial mediation.

OPEN IN VIEWER

Table 11.

Mediation, indirect effects, and total effects.

H	Relationship	Effect type	Estimate	β	SE	z	p-value	95% CI	Decision
H6	ER → MOT → PERF	Indirect	0.109	0.106	0.043	2.531	0.011	[0.025, 0.193]	Supported
H7	IR → MOT → PERF	Indirect	0.197	0.172	0.060	3.268	0.001	[0.079, 0.314]	Supported
—	ER → PERF	Total	0.402	0.391	0.098	4.103	<.001	[0.210, 0.595]	Significant
—	IR → PERF	Total	0.382	0.334	0.101	3.790	<.001	[0.184, 0.579]	Significant

The indirect effect of intrinsic rewards on employee performance through motivation was also positive and statistically significant (β = 0.172, z = 3.268, p = 0.001, 95% CI [0.079, 0.314]), supporting H7. Although the direct effect of intrinsic rewards on employee performance was not statistically significant, the significant indirect effect indicates that intrinsic rewards are associated with employee performance mainly through motivation.

The total effects were statistically significant for both reward dimensions. The total effect of extrinsic rewards on employee performance was β = 0.391 (z = 4.103, p < .001, 95% CI [0.210, 0.595]), while the total effect of intrinsic rewards was β = 0.334 (z = 3.790, p < .001, 95% CI [0.184, 0.579]). These results show that both extrinsic and intrinsic rewards had significant total effects on employee performance when direct and indirect pathways were considered together.

The structural model provided support for six of the seven proposed hypotheses. The results confirmed the significant effects of extrinsic and intrinsic rewards on motivation, the significant effect of motivation on employee performance, and the significant direct effect of extrinsic rewards on employee performance. The direct effect of intrinsic rewards on employee performance was not statistically significant; however, its indirect effect through motivation was significant. Thus, the findings indicate that motivation plays an important mediating role in linking reward practices with employee performance.