A Cross-Level Analysis of Linking Transformational Leadership and Employee Creativity to Enhance Task Performance: The Mediating Effects of Person-Supervisor Fit and Psychological Empowerment ()
1. Introduction
Creativity is generating or recognizing novel and useful ideas to solve problems and engage with a person, a group, or a society (Zhou & George, 2003) . Creativity involves both problem-solving and problem-finding (Sawyer & Henriksen, 2024) . Employee creativity has been recognized as critical for organizational growth and success (Tse, To, & Chiu, 2018) . Employees’ creativity is an important factor for organizational goals and sustainability competitive advantage (Suifan, Abdallah, & Al Janini, 2018) . Understanding organizational creativity is a high priority for research on organizational behavior (Fetrati, Hansen, & Akhavan, 2022) . Employee creativity is of great significance regarding organizational effectiveness and survival during the crisis (Chaudhuri, Grandhi, Vrontis, & Chatterjee, 2023) , organizational innovation (Nguyen, Hooi, & Avvari, 2023) , and firm performance (Mahmoud, Mustamil, & Seng, 2023) . Employee creativity is critical for an organization’s innovativeness and survival in today’s competitive business environment (Lua, Liu, & Shalley, 2023) . Developing effective leadership, such as transformational leadership, has been concerned with a useful way to motivate employees to generate novel and useful ideas for services, practices, and procedures (Al Harbi, Alarifi, & Mosbah, 2019; Tse et al., 2018) . Several studies have indicated that transformational leadership positively impacts creative outcomes (Tse et al., 2018) . Given the evidence for its effectiveness, researchers have begun to investigate different underlying mechanisms which enhance employees perception of their workplace creativity (i.e., Kasımoğlu & Ammari, 2020; Ma & Jiang, 2018; Ma, Jiang, Wang, & Xiong, 2020 ).
Despite past empirical research linking empowerment (i.e., transformational leadership and psychological empowerment) to creativity, three specific research analyses still need clarification. First, most recent streams of research on transformational leadership have been vague concerning the fact that these perspectives identify leadership at two levels of analysis, i.e., either at an individual level or a group/organizational level. For example, Gumusluoglu and Ilsev (2009) classified transformational leadership and creativity at two levels of analysis (i.e., the organizational level and the individual level). Carmeli, Gelbard, and Reiter-Palmon (2013) , McMahon and Ford (2013) , Eisenbeiß and Boerner (2013) , and Paulsen, Callan, and Saunders (2013) failed to classify transformational leadership or employee creativity at either an organizational level or an individual level of analysis. Second, the conflicting findings of previous studies regarding creativity at the individual level are of particular interest to the present research. For example, Shin and Zhou (2003) found that employees exhibit more creativity under transformational leadership. Transformational leadership and followers’ radical creativity from industry-university collaborative 293 team leaders-followers (Nabi, Liu, & Hasan, 2023) . Third, prior research on creativity has largely been guided by intrinsic motivation theory. Along this line, intrinsic motivation has been considered as a mediating variable through which contextual factors contribute to employee creativity (Oldham & Cummings, 1996) .
Transformational leadership augments employee creativity by applying the componential theory of creativity, social cognitive theory, and prosocial motivation theory (Chen, Li, & Tang, 2009; Gong, Huang, & Farh, 2009) . Indeed, transformational leadership has been treated as individual-level analysis to predict on employee creativity through mediating variable (i.e., creative self-efficacy) (i.e., Chaubey, Sahoo, & Khatri, 2019 ), work engagement (i.e., Fu et al., 2022 ), employee adaptiveness (i.e., Żywiołek, Tucmeanu, Tucmeanu, Isac, & Yousaf, 2022 ), and psychological empowerment and innovation climate (i.e., Koh, Lee, & Joshi, 2019 ). However, a cross-level analysis between organizational-level (or team level/group-level) and individual-level analysis is still limited for testing the relationship between transformational leadership and employee creativity (i.e., Dong, Bartol, Zhang, & Li, 2017; Kim, Park, & Kim, 2019; Miao & Cao, 2019; Wang, 2020 ). Based on the above research gap, this study treated transformational leadership at the organizational level (or team level/group level) to predict employee creativity at individual-level analysis. In this manuscript, person-supervisor fit and psychological empowerment have been treated as a mediating effect to predict an outcome variable of employee creativity.
By addressing the above-referenced research gaps, the present study draws upon multiple perspectives of social learning theory (Bandura, 1977) , leadership theory (transformational leadership) (Bass, 1985) , person-supervisor fit theory (Kristof-Brown, Zimmerman, & Johnson, 2005) , self-determination theory (Gagné & Deci, 2005) , and relevant psychological empowerment literature (Spreitzer, 1995) to focus on the following research objectives: Firstly, this study aims to examine the influence of transformational leadership on personsupervisor fit at the organizational level. Secondly, this study examines the influences of psychological empowerment on employee creativity and task performance at the individual level. Finally, at the cross-level of analysis, this study also investigates the influence of transformational leadership on psychological empowerment and the influences of person-supervisor fit on psychological empowerment and employee creativity. The cross-level mediating effects of person-supervisor fit and psychological empowerment are also examined, as shown in Figure 1.
This study brings contributions to the literature in two important ways. First, by examining psychological empowerment as a mediating variable, this study extends previous research frameworks that primarily have examined the direct or main effect of transformational leadership on psychological empowerment. The mediators of this relationship may provide theoretical insight into how transformational leadership influences employee creativity and task performance. Second, this study operationalizes the stream of research on the impact
Figure 1. Mediating effect of person-supervisor fit and psychological empowerment.
of person-supervisor fit on employee task performance through a mediating effect of psychological empowerment. This study contributes to the literature by explaining how transformational leadership and person-supervisor fit relate to employee creativity and task performance through a cross-level mediation of psychological empowerment and person-supervisor fit. Based on the above discussions, the following research framework is developed, as shown in Figure 1.
2. Literature Review
2.1. Transformational Leadership and Person-Supervisor (P-S) Fit
Leadership is defined as the capability of an individual to exercise influence and exert control over other members to help a group or organization achieve a vision or set of goals (Robbins & Judge, 2013; Yukl, 2010) . According to leadership theory, transformational leadership influences followers through a special leader-follower relationship (Conger, 1999) . Most of the extensions of literature on leadership integrate four sub-dimensions of transformational leadership (i.e., inspirational motivation, intellectual stimulation, supportive leadership, and personal recognition) (Wang & Howell, 2010) to predict person-supervisor fit. Leadership is an influential relationship between leaders and followers to achieve an organizational goal or objective (Veasna, 2013) . Transformational leadership attributes are key determinants influencing the supervisor-employee relationship. If supervisors and employees feel that their values match well, this may lead them to feel more satisfied with their jobs and work environment (Van Vianen, Shen, & Chuang, 2011) . Both organizations and leaders share the same value that causes them to resist changes, so high levels of P-S fit may be necessarily motivated by transformational leader behavior (Guay, 2013) . Transformational leadership significantly impacts employees’ person-job and person-supervisor fit in China (Bui, Zeng, & Higgs, 2017) . The relationship between transformational leadership and employee individual values is based on the person-supervisor fit theory (Zhou, Zhang, Lyu, & Zhang, 2018) . Indeed, the P-E fit theory proved that leadership styles (i.e., servant leadership) can enhance person-group fit (P-G fit) and person-supervisor fit (P-S fit) (Safavi & Bouzari, 2020) . The cross-level interaction between leaders and subordinates’ influences person-supervisor (P-S) fit perceptions in 10 companies in the United States (Guay, Kim, Oh, & Vogel, 2019) . Thus, this study argues that transformational leadership is one of the most important ways to improve employees’ person-supervisor fit in the workplace. Based on the above rationale arguments, the following hypothesis is proposed:
Hypothesis 1: Transformational leadership has a positive influence on person-supervisor fit.
2.2. Transformational Leadership and Psychological Empowerment
Transformational leaders also enhance follower empowerment by providing meaning and challenge to their work (Avolio, Zhu, Koh, & Bhatia, 2004) . Transformational leaders empower subordinates through enhancing subordinates’ psychological empowerment regarding four facets: impact (i.e., the degree to which employees feel their work affects their organization), competence (i.e., perceived ability to accomplish work-related tasks), meaning (i.e., intrinsically caring about work tasks), and self-determination (i.e., perceived self-determination or autonomy at work) (Zhang & Bartol, 2010) . Indeed, the relationship between transformational leadership and psychological empowerment can also be explained by the social learning theory (Zohar & Tenne-Gazit, 2008) . In terms of the degree to which cross-level analysis is used, this study argued that transformational leadership plays an important role in increasing the psychological empowerment of followers (Joo & Lim, 2013) . According to Fong and Snape (2013) , in both unit-level and individual-level analysis, leadership style (i.e., empowering leadership) and individual outcomes, empowering leadership was associated with psychological empowerment at both levels. In leadership styles (i.e., empowerment leadership) are positively related to subordinates’ perception of psychological empowerment (Amundsen & Martinsen, 2015) . Leadership styles (i.e., transactional and transformational leadership) are among the organizational factors suggested to facilitate and enhance employee psychological empowerment (Kim & Shin, 2017) . Leadership styles (i.e., empowering, transformational, servant, and transactional leadership) ranked high among the best predictors of employees’ psychological empowerment (Schermuly, Creon, Gerlach, Graßmann, & Koch, 2022) . Based on the above research arguments, the following hypothesis is proposed:
Hypothesis 2: Both (a) individual-level and (b) group-level of transformational leadership will be positively related to psychological empowerment.
2.3. Person-Supervisor Fit and Psychological Empowerment
Person-supervisor (P-S) fit is broadly defined as the perceived fit between employee and supervisor characteristics (i.e., values, personality, and behavioral styles) (Kristof-Brown et al., 2005) . The concept of psychological empowerment is likely to be impacted by the degree to which the supervisor-employee dyad shares the organization’s values and fits the organization (Gregory, Albritton, & Osmonbekov, 2010) . Ballout (2007) argued that P-S fit is positively associated with psychological empowerment. P-S fit may increase subordinates’ perceptions of empowerment practices aimed at decision-making that enable them to carry out their tasks (Subramony, 2009) . In person-organization fit theory suggests that person-supervisor fit supports employee psychological empowerment to improve work performance (Lau, McLean, Hsu, & Lien, 2017) . Previous studies have indicated that person-supervisor fit affects innovative work behavior and psychological empowerment (Melina & Sandroto, 2020) . Based on the above research argument, the following hypothesis is proposed:
Hypothesis 3: Both (a) individual-level and (b) group-level of relationship between person-supervisor fit psychological empowerment.
2.4. Psychological Empowerment and Employee Creativity
Empowered employees are more likely to exhibit creative (Jung, Chow, & Wu, 2003) . Zhou (2003) also argued that people must be empowered before developing intrinsic motivation, promoting employee creativity at work. Employee creativity requires a workforce that is high on psychological empowerment and role job satisfaction (Sangar & Rangnekar, 2014) . Empirical studies have proven a positive impact of psychological empowerment on creativity. For instance, Speklé, van Elten, and Widener (2017) demonstrated that psychological empowerment positively influences creativity. Creativity and psychological empowerment concepts have become popular issues in today’s turbulent and competitive business environment (Zhang, Kim, & Ding, 2023) . A recent study investigates the relationship between psychological empowerment and employee creativity (Nguyen & Doan, 2023) . Psychological empowerment is key to employee creative performance (Sahadev et al., 2024) . Based on the above arguments, this study proposes the following hypothesis:
Hypothesis 4: Psychological empowerment positively influences employee creativity.
2.5. The Mediating Effect of Person-Supervisor Fit on the Relationship between Transformational Leadership and Employee Creativity
Leadership theory has been identified as an important factor contributing to the culture and climate of the organization and perceptions of support for individual creative behavior (Reiter-Palmon & Illies, 2004) such as employee creativity (Amabile, Schatzel, Moneta, & Kramer, 2004) . According to leadership theory, transformational leadership may directly encourage employee creativity (Jung, Wu, & Chow, 2008) , but it can also indirectly affect employee creativity through a mechanism of person-supervisor fit. This study urges that transformational leadership is positively associated with followers’ creativity (Al-Amin, 2017) . Another study indicated that transformational leadership greatly contributes to the linkage between organizational innovation at the corporate level and creativity at the individual level (Al Harbi et al., 2019) . Psychological empowerment mediates the relationship between benevolent leadership and employee creativity in the dyadic relationship of 344 subordinates and 132 leaders/supervisors from 32 SMEs in the manufacturing industry of Ghana (Gyamerah, He, Asante, Ampaw, & Gyamerah, 2022) . Transformational leadership can intrinsically motivate employees and play a crucial role in promoting individual employee creativity (Żywiołek et al., 2022) . Based on the above research arguments, the following hypothesis is proposed:
Hypothesis 5: Both (a) individual-level and (b) group-level of person-supervisor fit mediates the relationship between transformational leadership and employee creativity.
2.6. Person-Supervisor Fit and Employee Creativity
According to person-environment fit theory, employee have high perceived work environment will enhance the creative work carried out in organizations (Kristof-Brown et al., 2005) . The fundamentals of person-supervisor fit (P-S) create organizational values and individual employees’ perceptions of the working environment (Vilela, González, & Ferrín, 2008) . Indeed, P-S fit has been shown to potentially influence individual employee outcomes, such as job satisfaction, productivity, creativity, and stability that are compatible with the characteristics of the working environment or unique work styles (Vilela et al., 2008) . P-S fit provides a suitable theoretical perspective to investigate the congruence between persons and organizations in employee creativity (Saraç, Efil, & Eryilmaz, 2014) . P-S fit affects employee creativity by drawing on social exchange theory and identifying leader-member exchange (LMX) (Seong & Choi, 2019) . Drawing on social exchange theory and identifying leader-member exchange, the effect of P-S fit on the creativity of 167 employee-supervisor dyads was confirmed (Seong & Choi, 2019) . Thus, this study assumes that employees will be more creative when they perceive their values as matched with their supervisors’ characteristics and working environment. Therefore, the following hypothesis is proposed:
Hypothesis 6: Both (a) individual-level and (b) group-level of positive relationship between person-supervisor fit and employee creativity.
2.7. Psychological Empowerment and Task Performance
Empowerment is one of the most important factors driving organizational effectiveness and individual task performance (Ahearne, Mathieu, & Rapp, 2005) . The literature on empowerment has argued that psychological empowerment positively influences job performance (Chiang & Hsieh, 2012; Wallace, Johnson, Mathe, & Paul, 2011) . Psychological empowerment enhances an individual’s task performance among those engaged in an international experiential game-based learning project (Curran, Arroteia, Blesa, Musteen, & Ripollés, 2021) . Empirical evidence confirms the positive impact of employee psychological empowerment on task performance (Juyumaya, 2022) , but is partially mediated by work engagement. Psychological empowerment positively correlates with employee work performance (Liu & Ren, 2022) . Psychological empowerment is directly related to individual employee performance (Mahmoud, Ahmad, & Poespowidjojo, 2022) . Psychological empowerment influences individual employees and organizational outcomes, but there is still a need to examine how empowered employees achieve higher performance levels (Ochoa Pacheco, Coello-Montecel, & Tello, 2023) . Indeed, psychological empowerment positively influences job performance (Ochoa Pacheco & Coello-Montecel, 2023) . Based on the above research arguments, the following hypothesis is proposed:
Hypothesis 7: Psychological empowerment has a positive influence on task performance.
2.8. The Mediating Effect of Psychological Empowerment on the Relationship between Transformational Leadership and Task Performance
In line with previous empirical studies, transformational leadership is positively related to employee creativity and performance (Si & Wei, 2011) . Leadership theory suggests that transformational leaders motivate individuals to achieve higher levels of both individual and group performance (Wang, Oh, Courtright, & Colbert, 2011) . Along this vein, transformational leadership has been extensively investigated regarding its effects on employee task performance as developed via influences on the psychological empowerment of followers. In leadership research contexts of the individual level analysis, psychological empowerment can be treated as an independent variable affect employee motivational mechanism (i.e., Schermuly & Meyer, 2020 ) and employees’ voice behavior (i.e., Ilyas, Abid, Ashfaq, Ali, & Ali, 2021 ) the final dependent variable is affected transformational leadership (i.e., Pradhan, Panda, & Jena, 2017; Schermuly et al., 2022 ), the mediating effects of the relationship between transformational leadership and employee work attitudes (i.e., Lan & Chong, 2015 ), organizational commitment (Avolio et al., 2004) , innovative climate (i.e., Sagnak, Kuruoz, Polat, & Soylu, 2015 ), task performance and citizenship behavior (i.e., Dust, Resick, & Mawritz, 2013 ), career satisfaction (i.e., Joo & Lim, 2013 ), innovative work behavior (i.e., Grošelj, Černe, Penger, & Grah, 2021; Stanescu, Zbuchea, & Pinzaru, 2021 ), employee outcomes (i.e., Saira, Mansoor, & Ali, 2021 ), followers’ attitudes (i.e., Barroso Castro , Villegas Periñan, & Casillas Bueno, 2008), organizational identification (i.e., Bose, Patnaik, & Mohanty, 2021 ), employees’ voice behavior (i.e., Ilyas et al., 2021 ), project success (i.e., Fareed, Su, & Aslam, 2023 ), emotional labor (i.e., Cheng, Liu, & Zhou, 2023 ), psychological health (i.e., Tripathi & Bharadwaja, 2020 ), and the moderating variable between transformational leadership and organizational citizenship behavior (i.e., Jha, 2014 ).
Transformational leadership improves employee performance (Buil, Martínez, & Matute, 2019) . In the leadership context, the effects of leadership style (i.e., empowering leadership) on employees’ job performance through the mediation of psychological empowerment (Kundu, Kumar, & Gahlawat, 2019) . Another study identifies the effect of transformational and transactional leadership on task performance and the mediating role of psychological empowerment (Ambad, Kalimin, Ag Damit, & Andrew, 2021) . Indeed, managers’ transformational leadership drives employees’ psychological empowerment and, in turn, task performance (Guerrero, Chênevert, Vandenberghe, Tremblay, & Ben Ayed, 2018) . This study draws on social identity theory (SIT) and social exchange theory (SET) to explore the mediating role of the psychological empowerment relationship between transformational leadership and task performance. Based on the above research arguments, this study treats the research variable of psychological empowerment as “the mediating effect” to the cross-level analysis of the research hypotheses already proposed in Figure 1. Thus, this study assumes that transformational leadership is positively related to psychological empowerment, further improving employee task performance. Based on the above research arguments, the following hypothesis is proposed:
Hypothesis 8. Psychological empowerment positively mediates the cross-level relationship between transformational leadership and task performance.
2.9. The Mediating Effect of Psychological Empowerment on the Relationship between Person Supervisor Fit and Task Performance
In the extension literature on person-environment fit, it has been consistently argued that person-supervisor fit also increases performance by empowering employees to perform their jobs independently (Van Vianen et al., 2011) . Employees’ highly perceived psychological empowerment improves their job performance at the workplace (Huang, 2012) . This study argues that employee perceptions of psychological empowerment mediate the relationship between P-S fit and task performance. Prior research has identified the mediating effect of psychological empowerment on the relationship between person-organizational (P-O) fit and in-role employee performance (Afsar & Badir, 2016) , little is known about the mediating role of psychological empowerment on the influence of P-O fit on employees’ task performance. The relationship between employees’ perceived P-O fit and employee-rated job performance depends on the perceived P-O fit of the supervisors (Hamstra, Van Vianen, & Koen, 2019) . In line with their research arguments, this study proposes that the P-S fit and employee’s task performance linkage is mediated by employee perceived psychological empowerment, then the following hypothesis is proposed:
Hypothesis 9. Psychological empowerment mediates the cross-level relationship between person-supervisor fit and task performance.
2.10. Employee Creativity and Task Performance
Employee creativity is the individual ability to generate creative approaches to solving problems more imaginatively (Pearsall, Ellis, & Evans, 2008) , which improves their job performance (Gong et al., 2009) . Employee creativity is generally believed to benefit a firm’s performance (Dul & Ceylan, 2014) . There is a significant positive effect on employee creativity and performance (Ximenes, Supartha, Manuati Dewi, & Sintaasih, 2019) . Firm performance was predicted by employee creativity (Yamin, 2020) . Employee creativity and sustainable organizational performance are stronger within firms that have employees with high learning capabilities (Muñoz-Pascual & Galende, 2020) . Empirical studies also indicate that engagement is related to individual employee performance (Pattnaik & Sahoo, 2021) . Employee creativity enhances Malaysian SMEs’ and job performance (Ismail et al., 2021) . Based on the above research arguments, the following hypothesis is proposed:
Hypothesis 10. Employee creativity positively influences employee’s task performance.
3. Methodology
3.1. Study Sites
This study also focuses on garment factory subsidiaries of Taiwan region and the Chinese mainland in Cambodia. Among China, Vietnam, Bangladesh, and Cambodia, Cambodia is considered one of the most attractive destinations for foreign investments in the garment industry in terms of its large pool of cheap labor. Cambodian textiles and garment exports are among the key reasons for the growth of the country’s economy, and the maximum GDP increase is due to this sector (Asuyama & Neou, 2012) . Cambodia’s garment sector is an essential indicator of economic growth, with the proceeds of the garment factories representing about 15% of Cambodia’s total GDP. It was responsible for about 50% of manufacturing employment (McKay & McKenzie, 2020) . The sector accounted for 11% of the economic growth. It contributed around 50% of Cambodia’s real GDP growth in 2021, with 67% of the country’s total merchandise exports to Europe and the US (International Labour Organization, 2021) and almost 700 factories across the country using around 671,509 workers. Companies in Taiwan region and the Chinese mainland remains the leading source of foreign investment in Cambodia’s Garment, Footwear, and Travel Goods (GFT) sector, accounting for 66.3% of approved new investments in 2021 (International Labour Organization, 2021) . Garment factory subsidiaries of Taiwan region and the Chinese mainland in Cambodia is chosen as the study sample site because it is expected that foreign companies can transform their leadership styles (i.e., transformational leadership) to enhance the working relationship between leaders and subordinates in the same work units and can build work efficiency through the creativity of individual subordinates.
3.2. Sampling Procedures
In a cross-sectional study of the dyadic relationship between leaders and subordinates, as Walumbwa et al. (2010) recommended, each manager or leader provided ratings for at least three subordinates. Then, this study collects data from the dyadic relationship between leaders and employees who work in the Research and development (R&D) departments in each garment factory invited to participate in the survey. A cross-sectional study (i.e., Zikmund, Babin, Carr, & Griffin, 2013 ) is adopted to select Chinese garment factories in Cambodia. A purposive sampling technique (i.e., quota sampling) (Cooper & Schindler, 2014) is also adopted to select 1:3 dyadic relationships from 1 leader and three subordinates from each R&D department. A self-administered survey is used to distribute the questionnaires to the HR department. Therefore, the validated 112 sample sizes from leaders and 336 subordinates from 112 garment factories owned by Chinese investment in Cambodia. Then these final sample sizes were appropriated for Hierarchical Linear Modeling (HLM) analysis (i.e., Raudenbush, Bryk, Cheong, Chongdon, & du Toit, 2011 ) in this study.
3.3. Measurement Scales
To minimize a common variance bias, as recommended by Podsakoff, MacKenzie, and Podsakoff (2012) , this study collected the data from three different sources. First, at the organizational level, individual employees in the sample rated their managers (i.e., transformational leadership and person-supervisor fit). This study then averaged the individual perceptions as 3 subordinates divided by 3. Second, individual employees rated psychological empowerment. Finally, managers rated the level of creativity and task performance of their three subordinates. The original items (English) were translated into Chinese (Chinese leaders) and Cambodian language (Cambodian subordinates) by following Brislin’s (1980) translation-back-translation procedure to validate the meanings of measurement items.
3.3.1. Transformational Leadership
Transformational leadership consists of four sub-dimensions (i.e., inspirational motivation, intellectual stimulation, supportive leadership, and personnel recognition) and 18 items of transformational leadership were adopted from Wang and Howell (2010) . A 5-point scale ranging from “1 = not at all to 5 = frequently, if not always” was used for measurement.
3.3.2. Person-Supervisor Fit
Person-supervisor fit consists of four items of person-supervisor fit were operationalized from Brown and Trevino (2006) in which employees were asked to rate the extent to which they matched their supervisors/managers using a 5-point Likert scale, ranging from 1 = strongly disagree to 5 = strongly agree.
3.3.3. Employee Creativity
Employee creativity has 13 items, which adopted from Zhang and Bartol (2010) . The leaders were asked to indicate the extent to which they felt the following statements applied to their employees, using a 5-point scale ranging from 1 = “not all characteristic” to 5 = “very characteristic”.
3.3.4. Task Performance
Task performance has four items of task performance were adopted from Chen and Aryee (2007) . Each manager was invited to rate 3 subordinates. This procedure is also consistent with Yun, Takeuchi, and Liu (2007) . A 5-point Likert scale ranging from 1 = “strongly disagree,” to 5 = “strongly agree” was used for measurement.
3.3.5. Psychological Empowerment
Psychological empowerment consists of four sub-dimensions (i.e., meaning, competence, self-determination, and impact) and 12 items of psychological empowerment were adopted from Spreitzer (1995) . These items were previously validated by Zhang and Bartol (2010) . A 5-point Likert scale was self-reported by individual employees, ranging from 1 = “strongly disagree,” to 5 = “strongly agree”.
3.4. Control Variables
Individual employees’ educational backgrounds and job tenure are related to creativity (Gumusluoglu & Ilsev, 2009) and performance (Shalley, Zhou, & Oldham, 2004) . Thus, these two control research variables were included in this study. It was expected that employee’s levels of education and job tenure would be found to be key information related to their work creativity and performance. Personal information for the respondents was also included as control variables, which consisted of five items: age, gender, educational background, occupation, and job seniority (work experience). Organizations may have different cultures or management styles to manage employees. According to Hofstede (1980) , national cultures can be described using four dimensions (i.e., individualism versus collectivism, large or small power distance, strong or weak uncertainty avoidance, and masculinity versus femininity).
Hofstede and Hofstede (2005) mentioned that the most relevant dimensions for leadership and organization are individualism-collectivism and power distance. The perceptions of fit between leadership behavior and the collectivist orientation of Chinese leaders will likely strengthen the positive direction of individual performance and organizational effectiveness (Bass & Riggio, 2006; Gumusluoglu & Ilsev, 2009) . Jung and Yammarino (2001) reported that the effects of leadership behavior (i.e., transformational leadership) are stronger among collectivists than individualists
Power distance is particularly relevant to authority relationships, for leaders are important authority figures in the workplace, and there is an inherent power differential in the leader-follower relationship (Liden, Erdogan, Wayne, & Sparrowe, 2006) . At the societal level, power distance has received considerable attention in cross-cultural leadership studies (Ng, Koh, Ang, Kennedy, & Chan, 2011) . Power distance refers to individually held values about power, hierarchy, prestige, conflict with authority, and social distance in the workplace (Anand, Vidyarthi, & Rolnicki, 2018) . It guides an individual’s feelings, thoughts, and behaviors related to power and status issues in the workplace and play an important role in understanding interactions between leaders and subordinates (Adamovic, 2023) .
The concepts of individualism and collectivism have stimulated intense psychological inquiry into the relationship between cultural values and various social behaviors (Tower, Kelly, & Richards, 1997) . Collectivism is a “social pattern of closely linked individuals who define themselves as interdependent members of a collective (e.g., family, coworkers), whereas individualism as a cultural pattern stresses individual autonomy and independence of the self” (Vandello & Cohen, 1999: p. 279) . Individualism as a “focus on rights above duties, a concern for oneself and immediate family, an emphasis on personal autonomy and self-fulfillment, and the basing of one’s identity on one’s personal accomplishments” (Oyserman, Coon, & Kemmelmeier, 2002: p. 4) . Thus, this study uses two cultural dimensions (i.e., collectivism and power distance) as control variables to explain the proposed framework in the cross-level analysis of two cultures between Cambodia and China. Three individualism-collectivism items are adopted from Eby and Dobbins (1997) . Eight items of power distance were adopted from Kirkman et al. (2009) . A 5-point Likert scale which “1 = strongly disagree to 5 = strongly agree” was used for measurement. According to Eby and Dobbins (1997) , individualism-collectivism has been designed as an individual level, and the subordinates were invited to rate leaders. According to Kirkman et al. (2009) , research items related to power distance have been treated at the individual level of analysis and are rated by subordinates.
3.5. Analytic Procedures
This study treats transformational leadership as an organizational-level variable consistent with Shamir, Zakay, Breinin, and Popper (1998) and Gumusluoglu and Ilsev (2009) . This study also operationalizes person-supervisor fit as an organizational-level variable. Therefore, subordinates’ transformational leadership and person-supervisor fit ratings are aggregated at the organizational level by averaging their values for each organization. The Intraclass Correlation Coefficients (ICCs—ICC1 and ICC2) technique is adopted to assess the interrater reliability of judgments provided by each department and organization. The term interrater reliability refers to the degree to which judges are “interchangeable,” which is to say, the extent to which judges “agree” on a set of “judgments” (James, Demaree, & Wolf, 1984: p. 86) . The ICC1 coefficient represents the proportion of variance in ratings at the individual level attributed to group membership, whereas the ICC2 coefficient represents the reliability of the group-level means (Bliese, 2000) . According to Mathieu, Gilson, and Ruddy (2006) , the minimum cut-off value for ICC1 is 0.12, and for ICC2, it is 0.60. One-way ANOVA is implemented to provide empirical justification for aggregating subordinate ratings for transformational leadership and person-supervisor fit. The results show that between-group differences are significantly higher than within-group differences (transformational leadership: F = 5.12, p < 0.001 and person-supervisor fit: F = 4.87, p < 0.001). The ICC1 for transformational leadership is 0.246, the person-supervisor fit is 0.213; the ICC2 for transformational leadership is 0.67, and the person-supervisor fit is 0.68. The within-group agreement (rwg(j)) is also calculated at an organizational level of analysis. In the case of the 112 cross-samples, their rwg(j) mean is 0.82 for transformational leadership and 0.84 for person-supervisor fit. All of the means rwg(j) is greater than the conventionally accepted value of 0.70 (James, Demaree, & Wolf, 1993) . Then, these results show that the aggregation is appropriate and acceptable for this study.
4. Results
4.1. Reliability Test
AMOS 25 is used to produce the results of confirmatory factor analysis (CFA) to evaluate the goodness-of-fit model assessment for the measurement model for this study. Anderson and Gerbing’s (1988) procedure was adopted to assess the reliability and construct validity of the measurement model for this study. According to Koufteros, Babbar, and Kaighobadi (2009) , the CFA procedure consists of two-factor models: a first-order and a second-order-factor model. A first-order-factor model is adopted to examine the individual research constructs of both levels of analyses (i.e., organizational and individual levels). The results of this procedure indicate that the standardized loading for all items exceeds 0.70 and that the t-values are higher than |1.96| (p < 0.001), thus satisfying the threshold recommended by Hair et al. (2010) and Kline (2011) . Then, a second-order factor model is conducted to examine the overall model fit of each research construct at both levels of analysis.
The following goodness of fit indices are chosen for this analysis based on suggestions found in previous studies (Baumgartner & Homburg, 1996; Koufteros et al., 2009; Shumacker & Lomax, 2004) . Absolute fit indices are intended to assess the overall model-to-data fit for structural and measurement models together (Chen, Curran, Bollen, Kirby, & Paxton, 2008; Vieira, 2011) : Chi-square goodness-of-fit test (χ2), the ratio of χ2 to degrees of freedom (χ2/df) < 3, root mean squared error of approximation (RMSEA) < 0.05, goodness-of-fit index (GFI) > 0.90, and adjusted goodness-of-fit index (AGFI) > 0.90; comparative fit index (CFI) > 0.90, and non-normed fit index (NNFI) > 0.90. Indeed, parsimony-adjusted fit measures attempt to compensate for the complexity of models (Shumacker & Lomax, 2004) . These measures reduce the overall sizes of the measures of fit by a constant known as the parsimony ratio (PRATIO) (Reinard, 2006) .
The results (Table 1) of the overall CFA model of the organizational level of analysis showed that absolute fit indices (i.e., χ2 = 25.247; df = 17; χ2/df = 1.485; GFI = 0.952; AGFI = 0.899; RMR = 0.031; RMSEA = 0.066) and incremental fit indices (NNFI = 0.962; CFI = 0.987) satisfy the threshold recommended by previous researchers. Parsimony-adjusted measures indicate that the PRATIO is 0.607, the PNFI is 0.584, and the PCFI is 0.599.
The results (Table 2) of the overall CFA model of the individual-level of analysis satisfy the threshold as recommended by previous researchers (i.e., χ2 = 149.551; df = 62; χ2/df = 2.412; GFI = 0.944; AGFI = 0.905; NFI = 0.955; CFI = 0.973; RMR = 0.036; RMSEA = 0.065). Parsimony-adjusted measures indicated that the PRATIO is 0.681, the PNFI is 0.651, and the PCFI is 0.663. “There are no hard-and-fast rules for interpreting these coefficients; the closer they are to 1.0, the stronger the model fit is claimed to be” (Reinard, 2006: p. 447) .
Table 1. The results of the overall CFA model (organizational-level—n = 112).
Note: ***p < 0.001, and they are significant at a t-value > 1.96. A regression weight was fixed at 1.00. (†): AVE=average variance extracted was computed from a second order factor model. (‡): AVE was computed from the overall CFA model. The AVE formula was adopted from Hair et al. (2010: p. 777) . TFLPR = Mean score of Personal Recognition; TFLSL = Mean score of Supportive Leadership; TFLIS = Mean score of Intellectual Stimulation; TFLIM = Mean score of Inspirational Motivation. λ = Standardized loading.
Table 2. The results of the overall CFA model (individual-level—n = 336).
Note: ***p < 0.001, and they are significant at a t-value > 1.96. A regression weight was fixed at 1.00. (†): AVE = average variance extracted was computed from a second order factor model. (‡): AVE was computed from the overall CFA model. AVE was adopted from Hair et al. (2010: p. 777) . λ = Standardized loading.
4.2. Discriminant Validity
The data is collected from two sources (i.e., managers and employees). This procedure is expected to reduce common method bias (Podsakoff et al., 2012) . However, self-reported scales of psychological empowerment still present a possibility that common method variance might be a concern (MacKenzie & Podsakoff, 2012) . Its discriminant validity is tested in three steps to identify the potential impact of common method bias in this study (Lee, Veasna, & Sukoco, 2013) . First, a Harman one-factor test is conducted, which loads all the variables into a principal component factor analysis (Podsakoff, MacKenzie, Lee, & Podsakoff, 2003) . The results reveal that no single factor dominates; the analysis generated four factors, accounting for 78.65% of the total variance (the first factor only accounted for 21.46%). Second, the percentage of variance extracted for any two factors is compared with the square of the correlation estimate between them (Fornell & Larcker, 1981) . Table 3 and Table 4 report the inter-factor correlations and their squared values. Each variance-extracted estimate satisfies the requirements, as they are greater than the corresponding inter-factor squared correlation estimation. Finally, the χ2 difference test is used to examine the
Table 3. The correlation matrix among research variables of organizational-level.
Note: N = 112; **Correlation is significant at the 0.01 level (2-tailed). *Correlation is significant at the 0.05 level (2-tailed). Square root of Average Variance Extracted (AVE) appears as bold numbers along the diagonal. N/A = Not Available.
Table 4. The correlation matrix among research variables of individual-level.
Note: N = 336; **Correlation is significant at the 0.01 level (2-tailed). *Correlation is significant at the 0.05 level (2-tailed). Square root of Average Variance Extracted (AVE) appears as bold numbers along the diagonal. N/A = Not Available. Psychological empowerment (i.e., Meaning, Competence, Self-determinant, and Impact).
distinctiveness of each pair of factors with correlation values above 0.50 using the common method factor (Conger, Kanungo, & Menon, 2000) . All cases result in significant differences, confirming that the pairs are not collinear (Anderson & Gerbing, 1988) . Overall, the results of these tests give us confidence that there is strong discriminant validity among the research factors. As shown in Table 3 and Table 4, the correlations among the research factors are less than 0.50, which suggests that common method bias is unlikely to be a problem (Gefen, Straub, & Boudreau, 2000) .
4.3. The Results of Hierarchical Regression Analysis and Hierarchical Linear Modeling
This study employs a cross-level mediational framework between the organizational and individual levels of analysis. Four mediating procedure conditions are adopted to analyze the multilevel organizational data to test the cross-level mediation effects of Hypotheses 5, 8, and 9 (i.e., Mathieu & Taylor, 2007 ). This study also uses hierarchical linear modeling (HLM 7) to test the cross-level relationships (Hofmann, Giffin, & Gavin, 2000; Raudenbush et al., 2011) of Hypotheses 2, 3, and 6. Then, hierarchical regression analysis (SPSS 25 software) is adopted to test hypotheses 1, 4, 7, and 10. The threshold cut-off values, as recommended by Hair et al. (2010) , are assumed to evaluate the hypotheses testing, such as R2 > 0.10; F-value ≥ 4; (p < 0.05), and significant at a t-value > |1.96|.
At the organizational level of analysis (Table 5), a hierarchical regression analysis is used to test Hypothesis H1. The findings show that transformational leadership significantly impacts person-supervisor fit (β = 0.367, t-value = 4.875, R2 = 0.292, p < 0.001), which supported Hypothesis H1. Indeed, managers’ age and working experience significantly contribute to transformational leadership’s influences on person-supervisor fit. This study argues that managers with more work experience will have better congruence with their work environments in terms of matches between person-supervisor fit and leadership styles in the workplace. At the individual level, Table 5 shows that psychological empowerment is significantly related to employee creativity (β = 0.237, t = 4.766, R2 = 0.802, p < 0.001) and task performance (β = 0.648, t = 14.593, R2 = 0.607, p < 0.001), which supported for Hypotheses H4 and H7, respectively. Employee creativity is also significantly related to employee task performance (β = 0.369, t = 6.986, R2 = 0.498, p < 0.05), supporting Hypothesis H10.
5. Discussion
At the organizational level of analysis, the research findings indicate the effect of transformational leadership on person-supervisor fit to be significant (Hypothesis 1), which suggests that this could be a novel finding of this study. At the individual level of analysis, the results confirmed the effect of psychological empowerment on employee creativity to be significant (Hypothesis 4), and this is consistent with previous theoretical arguments, employee creativity increases when employees feel empowered, which have suggested that psychological empowerment makes a critical contribution to employee creativity (Imam, Naqvi, Naqvi, & Chambel, 2020) . According to theories of psychological empowerment, self-determination, and composition theory of creativity to examine the relationship between psychological empowerment and employee creativity (Nguyen & Doan, 2023) . Indeed, the findings also confirmed that psychological empowerment is positively related to task performance (Hypothesis 7). This finding is consistent with the previous research arguments that proposed that employees who can attain high levels of performance may rely on leaders to give them more
Table 5. The results of hierarchical regression analysis of organizational and individual-level.
Note: ***p < 0.001, **p < 0.01, *p < 0.05. They are significant at a t-value > |1.96|. A stepwise method was used.
authority or empowerment to handle their work effectively (Kundu et al., 2019) . Thus, this study argued that task performance of teachers will be more pronounced when employees are empowered (Ahmed & Malik, 2019) .
The results also showed that employee creativity is partially and significantly related to task performance (Hypothesis 10). This finding aligns with previous research arguments that have proposed that employee creativity is related to developing new ideas and approaches that improve individual task performance (Hirst, Knippenberg, & Zhou, 2009) and overall job performance (Semedo, Coelho, & Ribeiro, 2016) . According to the theory of individual creativity, employee creativity is increasingly becoming a key determinant of employee job performance (Wang, Huang, Davison, & Yang, 2021) . Indeed, employee creativity enhances innovative work performance (El-Kassar, Dagher, Lythreatis, & Azakir, 2022) .
At the cross-level perspective of relationship analysis, the research findings indicated that transformational leadership has a positive influence on psychological empowerment (Hypothesis 2), which is consistent with previous studies of Richardson and Vandenberg (2005) , and Schneider and George (2011) , who proposed that considerable empirical evidence should exist on an impact of transformational leadership on psychological empowerment. This result also aligns with Gumusluoglu and Ilsev (2009) , who reported that transformational leadership significantly affects psychological empowerment. According to conferring to the leader-member exchange (LMX) theory and social exchange theory, leadership styles (i.e., transformational leadership and servant leadership) is able to enhance the level of psychological empowerment among frontline employees in the hospitality context (Safavi & Bouzari, 2020) . The research findings also indicated that person-supervisor fit significantly impacts psychological empowerment (Hypothesis 3). This finding is partially consistent with Gregory et al. (2010) and Ballout (2007) , who proposed that person-environment fit positively affects employees’ psychological empowerment. Indeed, person-supervisor fit plays an important role in enhancing employees ‘psychological empowerment (Tan & Wu, 2021) .
Based on the research findings of this study, person-supervisor fit has no significant influence on employee creativity (Hypothesis 6). This study posited that psychological empowerment plays an important role in mediating the relationship between person-supervisor fit and employee creativity. The mediating effect of employee’s psychological empowerment on the relationship between person-supervisor fit and employee creativity was confirmed by Sobel’s test (i.e., Iacobucci, 2012; Sobel, 1982 ), in which the z-test = 2.357 > t-value = |1.96| and p = 0.018 < 0.05.
Regarding the cross-mediation effect, the results indicate that person-supervisor fit does not mediate the relationship between transformational leadership and employee creativity (Hypothesis 5). Similarly research idea, this research finding is inconsistent the results of Safavi and Bouzari (2020) . The research findings of this study showed that the key independent variable of transformational leadership directly affects employee creativity (Kasımoğlu & Ammari, 2020) . Indeed, this study was expected to suggest that the control variables of age, individual-collectivism, and power distance strongly influence the relationships between person-supervisor fit and employee creativity. Thus, it was assumed that person-supervisor fit is critical in mediating the relationship between transformational leadership and employee creativity. The research findings also indicated that psychological empowerment fully mediated the relationship between transformational leadership and task performance (Hypothesis 8), which is consistent with George and Jones (2012) , who proposed that the effects of such leadership styles (i.e., transformational leadership) on employee job performance and commitment depends on employees’ sense of empowerment. It has also been argued that employees will be more engaged in job creativity and performance if their managers or supervisors empower them to make decisions.
However, psychological empowerment was not found to mediate the relationship between person-supervisor fit and employee task performance (Hypothesis 9). This hypothesis, as proposed in this study, lacks empirical support. Another perspective is that the dyadic relationship between leaders and subordinates may produce a mismatch in the perceptions of the perceived fit between employee and supervisor characteristics (i.e., values, personality, and behavioral styles) (Guay, 2013) . From this perspective, the person-supervisor fit may contribute less to predicting task performance and psychological empowerment, making psychological empowerment less significant to employee task performance. Furthermore, power distance and individual collectivism significantly affect employee task performance (Table 6). Therefore, in this study, it is assumed that power distance and individual collectivism may play an important role as moderating variables to explain the relationships among person-supervisor fit, psychological empowerment, and task performance, respectively. For example, power distance (between leaders and subordinates) can moderate the impact
Table 6. The results of HLM of the cross-level relationships.
Note: ***p < 0.001, **p < 0.05. They are significant at a t-value > |1.96|. γ = Intercept (standardized coefficient). R2 calculations were computed following Hofmann et al. (2000) and Hofmann, Morgeson, and Gerras (2003) .
of person-supervisor fit on psychological empowerment and task performance in such a way that person-supervisor fit relates more positively to psychological empowerment and task performance when power distance is higher rather than lower.
5.1. Managerial Implications and Contributions
This research is the first to explore transformational leadership and its effects on employee creativity and task performance by conducting a multiple-level analysis through the mechanism of the cross-level effects and mediation effects of employee psychological empowerment and person-supervisor fit in a cross-sectional study of Cambodian and Chinese contexts. The research findings suggest that in practical settings, transformational leaders are likely to implement empowerment practices and foster an autonomous work environment, which increases employee creativity and task performance through psychological empowerment. The results also suggest that transformational leadership and person-supervisor fit significantly affect employee creativity and task performance through the potential mediating process of psychological empowerment. Critically, person-supervisor fit (P-S fit), which exists in the dyadic relationship between individuals and their managers, is likely to impact employee motivation and organizational effectiveness (Lee, Reiche, & Song, 2010) . A supervisor or manager’s characteristics are the key determinant factors that influence employee behavior and attitudes (Van Vianen et al., 2011) . P-S fit is consistent with the concept that a socialization process provides new employees a framework for responding to their work environment and coordinating with other employees (Kim, Cable, & Kim, 2005) . Therefore, in this study, it was assumed that if employees feel that their values match those of their managers or supervisors, they will be more satisfied with their jobs and may also increase their work creativity. In turn, these employees are more likely to be committed to their organizations and are less likely to quit, which ensures that the organizations involved will receive greater returns on investments regarding recruitment, selection, and training (Kristof-Brown et al., 2005) .
The findings of this study should motivate managers to stimulate their subordinates by empowering them. Managers should understand that this mechanism could be a very effective way to enhance employee creativity and task performance. The findings suggested that adopting transformational leadership behavior is important to achieve higher levels of creativity and performance (Gumusluoglu & Ilsev, 2009) . The empirical findings of this study also provided evidence that transformational leadership and person-supervisor fit should be the subject of management training and employee development to help subordinates build confidence regarding performing tasks, providing a flow of challenging new ideas and encouraging subordinates to try new approaches to solving problems. This study also found that employees who feel empowered through their participation in decision-making will become more engaged in work creativity and task performance. In sum, drawing upon multiple theories of leadership, social learning, self-determination, and person-environment fit, this study indicates that psychological empowerment partially mediates the effects of the relationships between transformational leadership and person-supervisor fit on employee task performance respectively. The results highlighted the mechanisms of the impact of empowerment on employee creativity and task performance. They offered insights on enhancing employee creativity and task performance from psychological empowerment perspectives.
5.2. Limitations and Future Directions
Although this present study provides valuable insights into an understanding of the extension literature of transformational leadership, social learning, and social determination to explore employee creativity and task performance through the mechanisms of the cross-level and mediation effects of psychological empowerment and person-supervisor fit, there are a few limitations that should be recognized which may provide a departure for future research. First, the theoretical and empirical support is limited to the mediation effects of this research model. Also, it concerns the Chinese samples who work in 67 garment factory subsidiaries in Cambodia, with which this study focused on the dyadic relationship between leaders and individual subordinates. Second, the cross-sectional surveys of this study focused on the generational gaps of the dyadic relationship between leaders (i.e., organizational level of analysis) and subordinates (i.e., individual level of analysis).
Third, this study also focused on subordinate-leader pairs in each R & D department, which is a small portion of the total staff of such organizations. Therefore, future research should look closer at cross-functional units or teams than single units alone. This study proposed that cross-functional units (i.e., R&D, marketing, and HRM departments) will also be charged with developing and creating novel ideas to solve problems effectively. Thus, cross-functional units are brought together to perform unique tasks to build innovative services to achieve high organizational performance and customer service (Lussier & Achua, 2007) . Fourth, this study needed more empirical evidence and references to support the integration of two cross-sectional research studies. Industrial types, sample profiles, cultural traits, organizational cultures, and firm sizes may affect the results of this study. Thus, these variables need to be checked for their error term variances and should also be included in future studies. Therefore, future research must consider the selection of the same category of samples.
Fifth, previous studies have proposed that individual employees’ educational backgrounds and job seniority are related to creativity and individual performance (Gumusluoglu & Ilsev, 2009) . However, this study failed to prove that these control variables significantly relate to work creativity and performance. Therefore, organizational cultures, firm sizes, and job tenures may also impact the study results. Future research should replicate the findings in this study in other research contexts. Seventh, based on previous research, it is suspected that the emotional intelligence of leaders (Zhou & George, 2003) may play a critical role in enabling and supporting employee creativity and performance. Thus, these variables should be included among the organizational variables. Finally, achieving high levels of employee creativity through hiring and socialization is important in enhancing the organizational commitment necessary to meet competitive challenges (Kim et al., 2005) . It would also be interesting to examine how socialization tactics and transformational leadership jointly affect socialization outcomes beyond P-S fit, such as organizational citizenship behavior and performance.
The findings of this study indicated that person-supervisor fit was not significantly related to employee creativity (Hypothesis 6). This result is partially consistent with a research argument in a recent study suggesting that the relationship of personality facets between the dyadic supervisors-subordinates and employee creativity is not significant (Collins & Cooke, 2013) . Leader openness in matching the personality traits between managers and individual employees may play a critical factor in strengthening the relationship between person-supervisor fit and employee creativity. It is also expected that leader openness (Detert & Burris, 2007) may play an important role as a moderating variable in a cross-level effect that can improve the relationship between person-supervisor fit and employee creativity. For example, leader openness may moderate the impact of person-supervisor fit on employee creativity in such a way that person-supervisor fit relates more positively to employee creativity when individual employees perceive leader openness to be higher rather than lower. Furthermore, individual employees may have low perceptions of the fairness with which leaders may treat their subordinates. It is expected that procedural justice (i.e., organizational justice theory) (i.e., Ambrose & Schminke, 2009 ) might be treated as a moderating variable in a cross-level effect on the relationship between person-supervisor fit and employee creativity. Therefore, leader openness and procedural justice variables should be included in future studies.
Psychological empowerment does not significantly mediate the relationships between transformational leadership and employee creativity (Hypothesis 5) and person-supervisor fit and task performance (Hypothesis 9), respectively. This study argued that transformational leaders and person-supervisor fit can encourage subordinates to embark on self-development and self-discovery. This would result in higher levels of self-awareness among subordinates. Self-aware subordinates who actively embody the positive behaviors modeled by their leaders are more likely to have more in common with these leaders and share a collective social identity (Mhatre & Conger, 2011) . This is in line with social identity theory (SIT), which explains that individual employees tend to classify themselves into various social categories, such as organizational membership and age cohorts (Ashforth & Mael, 1989) . From the SIT aspect, individual employees and managers’ similarities in personality or lifestyle can be observed; they may belong to the same group. In contrast, they may be part of different groups if they have dissimilarities or are mismatched in terms of personality or lifestyle. Drawing from the above rationales, it is expected that a variable of leader-member similarly (Troster & van Knippenberg, 2012) can be treated as a cross-moderating effect to explain the relationships of transformational leadership and person-supervisor fit on employee creativity and performance.
6. Conclusion
The research findings of this study concluded that transformational leadership and person-supervisor are key determinant factors influencing individual employee creativity and task performance. This research finding is also consistent with the theoretical foundations proposed by previous researchers. The key positioning of this study was to test a mechanism of cross-level mediating effects of psychological empowerment and person-supervisor fit, as proposed in Hypotheses 5, 8, and 9. The research framework of this study consisted of double cross-level mediation effects of psychological empowerment and person-supervisor fit. Based on the research findings, Hypothesis 5 and Hypothesis 9 were rejected. This study argued that the shared explained variance of a mediation effect of psychological empowerment seems to be dominated by the explained variance of person-supervisor fit. Thus, it was concluded that only a cross-level mediation effect of psychological empowerment significantly contributes.
This study illustrated how the components of leadership style (i.e., transformational leadership) can help overcome some of the leadership challenges resulting from the differences between age cohorts. Based on the HLM results (see Tables 6-8), the age cohort of individual employees has a significant impact on the relationships of transformational leadership to employee creativity and task performance, respectively. Finally, this study also intended to explore how the generational gaps between leaders and subordinates affect individual behaviors at work regarding person-supervisor fit. The HLM results indicated that age cohort has a significant effect on the relationship of person-supervisor fit to psychological empowerment, employee creativity, and task performance, respectively. To validate the above statements, the results of Table 5 also confirm that the age cohort between leaders and subordinates does contribute to explaining the effects of the research variables in this study. It is argued that age gaps between age cohorts of leaders and subordinates may help in overcoming some of the leadership challenges in the research contexts of creativity and organizational innovation (Erickson, 2009; Erickson, Alsop, Nicholson, & Miller, 2009; Mhatre & Conger, 2011) . Thus, this study concludes that the age cohort between leaders and subordinates is critical in influencing the relationship between leaders and subordinates. This study also argues that organizations need to rely heavily on the skills of individual subordinates, who are certain to be in high demand (Mhatre & Conger, 2011) . According to the attribution theory, these employees will experience increased self-efficacy and job satisfaction (Robbins & Judge, 2013) .
Table 7. The results of HLM of the cross-level mediation effects (hypothesis 5).
Note: ***p < 0.001, **p < 0.05, *p < 0.01. They are significant at a t-value > |1.96|. γ = Intercept (standardized coefficient). R2 calculations were computed following Hofmann et al. (2000) and Hofmann et al. (2003) .
Table 8. The results of HLM of the cross-level mediation effects (hypotheses 8 and 9).
Note: ***p < 0.001 are significant at a t-value > |1.96|. γ = Intercept (standardized coefficient). R2 calculations were computed following Hofmann et al. (2000) and Hofmann et al. (2003) .
Note: ***p < 0.001; **p < 0.05 are significant at F-value ≥ 4.
Recently, company leaders have looked to the members of the age cohort of Generation Y as partners in success and job satisfaction (Erickson, 2009; Hershatter & Epstein, 2010) . Therefore, this study concludes that the results presented here contribute to our understanding of how the mechanisms of the effect of psychological empowerment and person-supervisor fit can manipulate individual employee creativity and offer insights on how to improve employee creativity from various transformational leadership perspectives. These research findings may also provide significant contributions to academics and professionals by which to understand matched pairs in employee-supervisor relationships to improve employees’ creative performance and meet demands for organizational innovation and expectations (refer to Table 9).