Employee agility’s moderating role on the link between employee vitality, digital literacy and transformational leadership with job performance: an empirical study

Abstract

This study examines the effects of employee vitality, digital literacy, and transformational leadership on job performance and the moderating role of employee agility. The result of this study indicates that the job performance of Human Resources practitioners is positively influenced by employee vitality, digital literacy and transformational leadership. Moreover, the findings also prove the moderating role of employee agility in these relationships. This research can be instrumental in further illuminating the determinants of job performance as well as the role of mediating factors. As the focus of this study is confined to Human Resources practitioners in manufacturing companies in northern Malaysia, and only considers employee agility as a mediating factor, future research is recommended to broaden its scope to encompass other roles within a company, explore other sectors, and consider other potential mediating constructs.

Keywords:

• Technological proficiency
• vitality assessment
• digital competence
• productivity enhancement
• HR
• innovations
• energy in workplace

Reviewing Editor:

• Huifen (Helen) Cai, Middlesex University Business School, United Kingdom

Subjects:

• Leadership
• Human Resource Management
• Organizational Studies

Subjects:

• Organizational Behavior
• Organizational Change
• Human Resource Management

1. Introduction

This study stems from the realization of the importance of HR practitioners (HRP) to organizations during a crisis as experienced during the recent COVID-19 pandemic. This has raised the question of the determinants of job performance (JP) of HRP so they can be better prepared when the next major work disruption hits. Hitherto, studies have been conducted on JP of white-collar workers who were affected by the disruptions caused by the pandemic (Bergefurt et al., 2022; Chinvararak et al., 2022; Wilms et al., 2022). Upon discovering and understanding how work changes brought about by COVID-19 impact the performance of HRP in Malaysia, considering the circumstantial evidence on the importance of HRP to organizations during tumultuous periods, and in preparation for a similar crisis to recur in the future, this study is conducted to determinant the influence of employee vitality (EV), digital literacy (DL) and transformational leadership (TL) on JP and the mediating role of employee agility (EA) in these relationship to enable organizations come up with more efficacious HR strategies to improve JP for HRP.

Having considered these, it is incumbent upon the researchers to empirically identify other factors that could positively influence JP of HRP to overcome the adverse effects of these work disruptions. Consequently, the outcome of this study can also be used to develop appropriate strategies to adapt human resources to the new reality to create effective interventions for dealing with their human resources in highly adverse and stressful situations in preparation for the next major work disruption. Furthermore, this study contributes to the theoretical field of Human Resources Management by identifying new determinants of JP as well as filling a notable gap by scrutinizing how EA moderates the relationship between the identified determinants and JP.

2. Theory

This study integrates the Work Adjustment Theory (WAT) (Dawis & Lofquist, 1984) and the Dynamic Capability Theory (DCT) (Teece, 2007) to provide insights into JP in changing environments. WAT posits that work is an interaction between individuals and their environment, which demands ongoing adaptation to ensure mutual satisfaction, also termed ‘work adjustment’. This continuous adaptation necessitates an agile workforce, characterized by continual learning, adaptability, and readiness for change (Elanthi & Dhanabhakyam, 2021; Storme et al., 2020). In the face of rapid changes, enhancing DL and vitality can foster an agile workforce, enabling HRPs to meet evolving work demands, as per the WAT. Hence, this study asserts that EA, influenced by DL and EV, can lead to improved JP.

Concurrently, one of the fundamental processes of DCT is ‘reconfiguration’ which demonstrates the organization’s ability to achieve adaptability in a changing environment (Bag et al., 2023). The theory, conceptualized as a triad of sensing, seizing, and reconfiguring (transformation), emphasizes the role of human capital in fostering agility and adaptability. Thus, cultivating an agile workforce can enhance an organization’s dynamic capabilities.

Given that workforce agility is a form of dynamic capability (Ajgaonkar et al., 2021), and considering that TL positively impacts dynamic capabilities (Lopez-Cabrales et al, 2017), this study also identifies TL as a potential factor that fosters EA.

3. Background and hypotheses development

3.1. Job performance

JP has various definitions across disciplines. It has evolved over time and become deeply intertwined with organizational practices. Initially viewed as a function of individual abilities and effort, as evident in Borman and Motowidlo’s (1993) framework of task and contextual performance, this perspective illuminated the multifaceted nature of JP. Over time, this understanding gradually morphed into seeing JP as a singular result of an employee’s work, with an emphasis on employer expectations.

In the 21st century, the definition was further broadened to include actions, behaviors, and outcomes contributing to organizational goals. This shift was captured by Koopmans et al. (2011), who proposed distinguishing productivity from performance, indicating that work performance involves more than simple goal achievement. Furthermore, performance appraisal processes often utilize diverse techniques to measure performance, focusing on job-related requirements, desired behaviors, and the avoidance of unacceptable behaviors. Later research further diversified this understanding by considering factors like employee behaviors, organizational contributions, and job task accomplishment.

More recently, Darvishmotevali and Ali (2020) opined that JP mirrors an employee’s capability to meet organizational expectations through their knowledge, skills, behavior, and ethical values.

Considering the impact of the pandemic on HRPs, this study defines JP in line with employer expectations, focusing on task performance and contextual performance as proposed by Koopmans et al. (2011).

3.1.1. Task performance

Researchers have defined task performance (TP) as the speed and efficiency with which an employee carries out assigned tasks (Darvishmotevali & Ali, 2020; Koopmans et al., 2014). This concept aligns with the job responsibilities outlined in the job description, making it a practical performance measure. This viewpoint is corroborated by Che et al. (2021), who described TP as the execution of officially recognized duties. Additionally, Luthfi et al. (2020) categorized TP into three types: routine TP, adaptive TP, and creative TP. As TP is inherently job-specific, it varies across different work roles and is often used as a metric for an employee’s proficiency in core technical activities.

3.1.2. Contextual performance

However, TP is just one dimension of JP, focusing primarily on task completion efficiency. The broader concept encompasses another dimension known as contextual performance (CP). This dimension represents extra-role behavior, such as aiding co-workers and supporting the organization, thereby contributing to the overall work environment (Ingusci et al., 2019; Yunarti et al., 2020). Unlike TP, CP involves behaviors not directly related to job tasks but have a significant impact on organizational, social, and psychological contexts (Widodo, 2021). It includes discretionary behaviors not explicitly required by any job yet contributing to the social context of all roles. Hence, both TP and CP independently contribute to an employee’s overall performance.

Motowidlo and Kell (2012) highlight CP as behavior that enhances organizational effectiveness by influencing the psychological, social, and organizational contexts of work. Unlike TP’s focus on work expectations, CP emphasizes behavioral aspects.

3.2. Employee vitality and employee agility

Vitality has been found to manifest itself in different ways e.g. high energy level, sense of serenity, healthy fatigue after a busy day and desire and readiness to return to work each day (Stoloff et al., 2019). Essentially, characteristics of vitality are also critical to good JP: feeling energized (Yang et al., 2017), demonstrating enthusiasm (Lantara, 2019), feeling alive (Gelbard et al., 2018), and resilience (Kašpárková et al., 2018). Nevertheless, as far as this study is able to verify, very few studies have explored the relationship between EV and JP.

This research proposes that EV has a positive influence on employee agility; an employee with vitality is likely to have more emotional energy, thus making them more proactive in the workplace (Jahanshahi et al, 2019). The relationship between EV and proactivity is also supported by other researchers such as Wörtler et al (2020) and de Jonge & Peeters (2019). Essentially, this proactivity relates to the positive energy and mental and physical strength possessed by an employee with high vitality.

In addition, the mental strength of an employee with high vitality may contribute to the resilience commonly found in them (Ungar & Theron, 2020; Whitfield & Wilby, 2021). Similarly, the connection between positivity – a characteristic of vital employees – and resilience is well-established (Magalinggam & Ramlee, 2021; Nadat & Jacobs, 2021; Ramli et al., 2021). Stengel (2018) noted that adults who practice positive thinking are more likely to persist with seemingly intractable problems. Positive thinking in individuals with vitality can help resolve problems and obstacles, contributing to a happier life (Thadchai et al., 2018) – all characteristics of a resilient person.

According to Kodden (2020), adaptability is a person’s ability, skill, disposition, willingness, and/or motivation to change or fit different tasks and social, or environmental features. This research also proposes that this willingness and motivation to change or fit different tasks and social or environmental features – which represent adaptability – are anchored in the positive mindset and energy of a person full of vitality (Duchek, 2019). As such, based on all the above, it is proposed that:

Hypothesis 1: Employee vitality has a positive relationship with employee agility.

3.3. Digital literacy and employee agility

Studies on DL have gained considerable traction since the late 20th century. It’s often equated with computer literacy, ICT literacy, information literacy, and media literacy. DL, focusing on the ability to access and reveal information through digital tools (Yildiz, 2020; (Canan Güngören et al., 2022), has become an essential skill that influences various aspects of personal and professional life (Mishra & Sharma, 2018; Rauth, 2023). According to Huu (2023), employees with greater digital autonomy are also more likely to engage in innovative work which leads to improved job performance.

Over time, the role of DL has extended to the HR field, with HR professionals expected to become digitally adept to align with the rising demand for data-driven, evidence-based HR practices that incorporate technology and design thinking principles (Rao, 2019; Styr, 2021).

Early researchers sought to define digital competence, identifying three primary dimensions: technological, cognitive, and ethical (Calvani et al., 2008). This framework was later expanded by Ng (2012a), who outlined three dimensions of DL comprising technical, cognitive, and social-emotional dimensions, emphasizing abilities ranging from technical proficiency and information synthesis to responsible internet usage and ‘netiquette’. Over time, other dimensions were gradually incorporated to further define the term DL, taking a more holistic view of the skill. For this study, DL will adopt the definition by Ng (2012a).

This study suggests that DL is key in facilitating organizational learning. The challenge lies in utilizing technology to simplify learning and make knowledge more accessible. In this regard, Bala et al. (2019) posited that IT utilization can enhance EA by fostering knowledge exchange and collaboration. Considering the dimensions of agility as proposed by Sherehiy combined with works by Alavi et al (2014) on enablers of EA, as well as understandings from the works of Muduli (2016), it is proposed that:

Hypothesis 2: Digital literacy has a positive relationship with employee agility.

3.4. Transformational leadership and employee agility

TL, first proposed by Burns (1978) and further elaborated by Bass (1999), is an interactive process that elevates the motivation and morality of both leaders and followers. TL, which emphasizes four key aspects—charisma, inspirational motivation, intellectual stimulation, and individualized consideration—has been found to enhance JP among other outcomes (Lai et al., 2020). Uniquely, TL also acts as a powerful catalyst for change (Ismaniah et al., 2020; Londoño-Proaño, 2021). Thus, its relevance in performance enhancement may lie in its ability to invoke change in individuals’ abilities, skills, and motivation.

A study on IT firms found that top management’s TL enhances workforce agility, impacting corporate reputation (Das, Mukhopadhyay & Suar, 2023). Additionally, exposure to TL was also found to boost organizational creativity, which could influence employee agility, as agile workers are typically more creative (Tessarini Junior & Saltorato, 2021).

Transformational leaders can also stimulate employees’ intrinsic motivation and promote their capacity for self-regulation – which are key elements of agility – enabling them to foster a climate of trust and psychological safety, which encourages employees to take risks, experiment with new ideas, and learn from failures—all essential elements for agility (Sasangohar et al., 2020). Furthermore, transformational leaders can facilitate the development of employees’ dynamic capabilities, such as sensing, seizing, and reconfiguring, which are critical for agility in the face of rapid changes and uncertainties in the business environment (Teece et al., 2016).

Building on past studies, this research suggests that leadership affects EA through the support and relationships it provides to employees. Considering the well-established relationship between TL and the enablers of EA such as relationship and support, job autonomy, and support and trust from management, it is proposed that:

Hypothesis 3: Transformational leadership has a positive relationship with employee agility.

3.5. Employee agility and job performance

In the rapidly changing landscape of the 21st century, adaptability and agility have become essential for both personal and organizational survival. Sherehiy (2008) delineated EA into three elements: proactivity, adaptability, and resilience. Essentially, EA reflects an individual’s capacity to respond to and prosper from workplace changes (Müceldili et al., 2020).

The COVID-19 pandemic has underscored the importance of an agile workforce capable of responding promptly to change, thereby sustaining JP amidst potential stressors. In this context, knowledge and IT skills have been proven to be important factors in promoting various dimensions of EA (Lai et al., 2021; Müceldili et al., 2020; Muduli, 2016; Salmen & Festing, 2021). Although various studies have explored individual dimensions of EA and their separate impacts on JP, no research has directly investigated the relationship between EA as a construct and JP. The current research aims to fill this gap.

Despite its growing significance in modern research, EA remains underexplored, with numerous definitions and dimensions. Amid varying definitions of EA (Franco & Landini, 2022; Petermann & Zacher, 2021; Sherehiy, 2008, Sherehiy & Karwowski, 2014), the scale adopted in this present study for EA, derived from the work by Sherehiy (2008), is the more widely used one (Salmen & Festing, 2021).

Agile employees are proven to improve organizational productivity, flexibility and ability to withstand challenges thus having a positive impact on organizational performance (Makori et al., 2022). In the context of HR practitioners, Mehrajunnisa et al. (2021) commented on how important it is for HR practitioners to be agile, and dynamic and be able to align themselves with current business trends to ensure sustainability and profitability of the company they work for. Agile HR practitioners can also effectively leverage technology to streamline HR processes, enhance employee engagement, and improve organizational performance (Madakam et al, 2019). Moreover, the uncertainties and challenges posed by major events such as Brexit and the recent COVID-19 pandemic underscore the importance of agility as agile HR practitioners can proactively manage these challenges, adapt HR practices to changing legal and economic contexts, and ensure the continuity and effectiveness of HR operations (Ridgway, 2019). Similarly, agile employees, with their dynamism and aptitude for swift skill application, also offer substantial performance benefits (Heydarbeigi et al., 2021). Hence, it is believed that constructs such as EV, DL, and TL likely contribute to these agile qualities which potentially mediate their relationship with JP.

Given that a significant portion of HRP’s performance is evaluated based on their ability to support the organization, it is proposed that:

Hypothesis 4a: Employee agility has a positive relationship with task performance.

Hypothesis 4b: Employee agility has a positive relationship with contextual performance.

3.6. Mediating effects

The determinants of JP are driven by the quest for optimal employee performance and the fostering of a high-performance workforce. According to extant literature, they can be grouped into three categories: organizational, environmental, and personal factors. Organizational factors include teleworking modalities, training, leadership styles, and job characteristics, while environmental factors encompass the physical work setting, organizational culture, and business sector. Finally, personal factors consist of individual traits such as personality, emotional intelligence, learning propensity, and creativity. These facets provide a robust framework for exploring the influence of additional factors, including vitality, digital literacy, transformational leadership, and agility, on JP. This comprehensive understanding is crucial as it offers a holistic perspective on how intersecting factors contribute to overall JP within a business context.

Despite the significance of EV, its relationship with JP remains underexplored. Earlier, Binyamin and Brender-Ilan (2018) proposed that vitality could influence JP through proactivity, a conclusion supported by Wörtler et al. (2020). A study by Op den Kamp et al. (2020), whose study focused on proactive vitality management and its impact on performance also demonstrated an indirect relationship between EV and JP. These studies thus indicate an indirect effect of EV on JP, suggesting that further exploration is needed in this area.

Employees with high vitality are better equipped to handle JP constraints (Choi et al., 2020; Ji et al., 2021; Mariappanadar, 2020). This ability proved critical during the COVID-19 pandemic, with the challenges of health protocols, limited resources, and remote work impacting performance. Higher levels of vitality have also been found to lead to increased productivity (Jabeen et al., 2021) as they contribute to positivity, mental and physical robustness, and reduced stress at work (Tummers et al., 2018).

The pandemic led to a decrease in vitality, largely due to the transition to remote work (Nordbäck et al., 2021), which may have contributed to the drop in JP during this period (Kitagawa et al., 2021). Considering de Jonge & Peeters’ (2019) conclusion that employees with high vitality exhibit elevated levels of energy, resilience, and proactivity—key dimensions of EA—this study hypothesizes that EA may mediate the relationship between EV and JP. Hence, this study proposes that:

Hypothesis 5a: Employee agility mediates the relationship between employee vitality and task performance.

Hypothesis 5b: Employee agility mediates the relationship between employee vitality and CP.

Over time, numerous studies have been conducted on the effect of DL on JP. For instance, Guzmán-Ortiz et al. (2020) in their study on the impacts of digital transformation on JP in Peruvian insurance firms, concluded that elements like customer service experience and digital processes greatly improved performance. Similarly, Mohd Abas et al. (2019) analyzed the correlation between DL and employee performance in Malaysia’s oil and gas industry. Their findings showed strong relationships between DL in technological, organizational, and environmental contexts and performance, reinforcing DL’s contribution to performance. These studies and others are pivotal in proving the link between DL and JP.

Hitherto, studies have been conducted on the impact of DL on JP across various fields (Mohd Abas, Yahaya & Fee Din, 2019; Nikou, De Reuver & Mahboob Kanafi, 2022). Generally, these studies agreed on the impact of DL on JP. However, this relationship may be attributed to the ability of DL to improve job satisfaction, worker autonomy (Cijan et al., 2019), task visibility (Timonen & Vuori, 2018), and employee well-being (Cazan, 2020). This is because digitalization bolsters knowledge flows, enabling employees to access vast amounts of information, enhancing efficiency, fostering communication (Fischer et al., 2020), and simplifying data management (Hamakawa et al., 2021).

Given the rise in workplace digitalization, the correlation between DL and JP is evidently crucial. Hence, this study proposes that DL can enhance JP by fostering proactivity, adaptability, resilience, and ultimately, EA. Given the close relation between DL and JP – especially white-collar workers – and considering the dimensions of agility as proposed by Sherehiy combined with works by Alavi et al on enablers of agility, as well as understandings from the works of Muduli, it is proposed that:

Hypothesis 6a: Employee agility mediates the relationship between digital literacy and task performance.

Hypothesis 6b: Employee agility mediates the relationship between digital literacy and contextual performance.

Known for its ability to positively impact employees’ JP, TL stands apart from other styles of leadership. It involves followers in processes related to personal factors within the organization (Salim & Zakaria, 2021), and it focuses on the authentic empowerment of followers (Pradhan et al., 2018). Given that TL also influences constructs like employee creativity (Fatimah & Martdianty, 2020; Shafi et al., 2020), employee learning (Akdere & Egan, 2020), and knowledge sharing (Yadav et al., 2019; Yin et al., 2019), it’s important to note that it plays a significant role in employee learning and improvement. These are essential for adaptability in a changing work environment.

Given the transformative nature of this leadership style in promoting adaptability, this research aims to validate the hypothesis that TL contributes to JP through the mediating effect of employee agility.

Despite limited research on the link between TL and EA, existing literature suggests potential correlations between TL with the dimensions of agility: proactivity, adaptability, and resilience. TL can foster employee proactivity by mitigating psychological distress and enhancing intrinsic motivation (Yi et al., 2019). Moreover, a meta-analysis by Park and Park (2019) has identified decision-making autonomy, coworker and supervisor support, a climate for innovation, and a learning organization – all outcomes of TL – as antecedents of performance adaptability. A later study in Poland also proved how a learning culture supported by TL influences change adaptability (Kucharska & Rebelo, 2022) further suggesting the relationship between TL and EA.

As TL could be related to employee agility through the proven relationship between transformational leadership and readiness to change (Meria et al., 2022; Peng et al., 2020; Rumijati et al., 2022) – which is closely associated with employee agility (Doeze Jager et al., 2021; Park & Park, 2020) – there are reasons to investigate the novel idea that transformational leadership is related to job performance through employee agility. Hence, this study proposes that:

Hypothesis 7a: Employee agility mediates the relationship between transformational leadership and task performance.

Hypothesis 7b: Employee agility mediates the relationship between transformational leadership and contextual performance.

The research framework of this study is shown in Figure 1 below.

Figure 1. Research framework.

4. Methods

4.1. Sampling

This study focused on full-time HRPs employed in manufacturing companies within the northern region of Malaysia. A purposive sampling strategy was used to select participants, the profile of which is shown in Table A1.

4.2. Measurement

A total of 1216 emails were sent to HRPs of relevant companies seeking their participation in the study. Of these, 557 HRPs responded and were issued questionnaires consisting of 44 items measuring EV, DL, TL, EA and JP. The response rate was 60.3%, with 336 responses received. After excluding 36 responses for not meeting the inclusion criteria, the final sample size was 300 HRP, accounting for 53.8% of the questionnaires distributed, thus ensuring the study’s robustness.

Dependent variable (job performance): In this research, JP was divided into task and CP. HRP rated their own TP using a 6-item scale from Williams and Anderson (1991), Lynch et al. (1999), and adopted by Talukder et al. (2018), scored on a 7-point Likert scale. CP was also rated using a 4-item scale from the same sources and scored similarly on a 7-point Likert scale.

Mediating variable (employee agility): EA is treated as a higher-order construct comprising three dimensions: proactivity, adaptability and resilience. Items measuring EA were adapted from the scale originally developed by Alavi et al. (2014) but was later adopted by Pitafi et al. (2018). Responses to these twelve items were rated through a 7-point Likert scale ranging from ‘1’ ‘strongly disagree’ to ‘7’ ‘strongly agree’.

Independent variables (employee vitality, digital literacy and transformational leadership): EV was gauged in this research using five items from Kark and Carmeli (2009), scored on a 7-point Likert scale. DL is treated as a higher-order construct comprising three dimensions: technical, cognitive, and social-emotional. It was measured using ten items as adapted from Ng (2012b), also scored on a 7-point Likert scale. Finally, TL was measured through seven items adapted from Carless et al., (2000), using a similar 7-point Likert scale.

Common method variance: To mitigate common method variance, this study adopted the recommendations of Tehseen et al., (2017) besides reducing ambiguity in the questionnaire through a pilot test, and statistical control remedies like Harman’s single-factor test and Rönkkö and Ylitalo’s (2011) six-step marker variable approach. Additionally, marker variables using four questions were also used, which according to Chin et al. (2013), can remove up to 70% of common method variance. These questions were adopted from studies by Lichtenstein et al. (1993) and used by Hampson et al. (2021) to measure price consciousness. The items are listed below:

1. I go to extra effort to find lower prices.

2. I shop at more than one store to take advantage of low prices.

3. The money saved by finding low prices is worth the time and effort.

4. I compare prices of at least a few brands before I choose one.

Lastly, a full collinearity test was conducted following Kock (2015) to identify any issues with common method variance.

4.3. Statistical methods

Data analysis for this study was performed using SPSS version 29 and Smart Partial Least Squares (PLS) version 4.0. Initial checks for missing values, outliers, and normality were conducted, and descriptive statistics were calculated. Common method variance was also addressed through Harman’s single-factor test, the use of marker variables, and a full collinearity test.

Next, the measurement model was evaluated using SmartPLS 4.0 to ensure the reliability and validity of the constructs, which included examining factor loadings for item correlations. Higher-order constructs, namely DL and EA, were then assessed for validity based on their lower-order constructs.

The structural model was then evaluated, beginning with assessing collinearity issues and examining the significance of the relationships within the model. Hypotheses were tested and coefficients of determination (r²) were assessed, followed by examining effect size (f²) and the model’s predictive relevance using PLS Predictive analysis. Finally, hypotheses testing was conducted for H1, H2, H3, H4a and H4b in addition to mediation analysis for the remaining hypotheses: H5a, H5b, H6a, H6b, H7a and H7b based on statistical significances of the total effect, direct effect, and specific indirect effect between variables.

5. Results

5.1. Descriptive statistics and correlations

The data for this study was inspected for outliers using regression analysis in SPSS and applying Casewise Diagnostics for standardized residuals, as proposed by Tabachnick and Fidell (2013). Based on the results of Casewise Diagnostics and Cook’s distance, no significant outlier issues were found so no data was excluded from the dataset. Although the Partial Least Squares Structural Equation Modeling (PLS-SEM) used is nonparametric and doesn’t require normality of data, the data set was still tested for extreme non-normality due to the potential problems this might cause in assessing parameter significance (Civelek, 2018). Based on skewness and kurtosis measures of the data, it was found that the data generally within the acceptably normal ranges according to Yim and Byon (2020) i.e. ±2 range for skewness and ±5 range for kurtosis. The result of the normality analysis is also shown in Table 1 below.

Table 1. Normality analysis.

		TP	CP	EV	DL	TL	EA
N	Valid	300	300	300	300	300	300
	Missing	0	0	0	0	0	0
Mean		5.9657	5.8852	5.5523	4.9106	5.7918	5.6962
Std. Deviation		.68619	.73559	.92333	.91373	.73760	.67462
Variance		.471	.541	.853	.835	.544	.455
Skewness		−.622	−.435	−1.122	−.261	−.555	−.664
Std. Error of Skewness		.141	.141	.141	.141	.141	.141
Kurtosis		.374	−.236	2.771	.034	−.020	.878
Std. Error of Kurtosis		.281	.281	.281	.281	.281	.281

5.2. Common method variance

This research used Harman’s single-factor test, marker variables, and a full collinearity test, to confirm the absence of common method variance (CMV). Harman’s single-factor test (Moorthy et al., 2019) showed a total variance explained by a single factor at 38.4%, well below the 50% threshold, implying no CMV issues. The test results are shown in Table A2. Next, marker variables were used to allow a comparison of the baseline model and the marker model (as shown in Table 2). The result revealed no significant differences, with an unchanged r² for CP and a negligible 0.2% change for TP and EA. The statistical significance of all paths remained consistent, thereby further supporting the absence of CMV in this studyalso revealed no significant model differences, further supporting the absence of CMV.

Table 2. Comparison between baseline model and marker included model.

Relations	Without Marker Variable			With Marker Variable
Direct effects	Path coef	p-values	Remarks	Path coef	p-values	Remarks
DL -> CP	0.213	0.001	Supported	0.212	0.002	Supported
DL -> EA	0.385	0.000	Supported	0.381	0.000	Supported
DL -> TP	0.166	0.017	Supported	0.162	0.021	Supported
EA -> CP	0.318	0.000	Supported	0.317	0.000	Supported
EA -> TP	0.373	0.000	Supported	0.372	0.000	Supported
EV -> CP	0.256	0.000	Supported	0.255	0.000	Supported
EV -> EA	0.293	0.000	Supported	0.290	0.000	Supported
EV -> TP	0.235	0.000	Supported	0.233	0.000	Supported
TL -> CP	0.137	0.016	Supported	0.137	0.016	Supported
TL -> EA	0.184	0.002	Supported	0.184	0.002	Supported
TL -> TP	0.129	0.017	Supported	0.130	0.017	Supported
Indirect effect	Without MV	p-values	Remarks	With MV	p-values	Remarks
DL -> EA -> CP	0.122	0.007	Supported	0.121	0.007	Supported
DL -> EA -> TP	0.144	0.005	Supported	0.142	0.005	Supported
EV -> EA -> CP	0.093	0.001	Supported	0.092	0.001	Supported
EV -> EA -> TP	0.109	0.001	Supported	0.108	0.001	Supported
TL -> EA -> CP	0.058	0.023	Supported	0.058	0.023	Supported
TL -> EA -> TP	0.069	0.018	Supported	0.069	0.018	Supported
Change in Coefficient of Determination (R^2) - with and without Marker Variables
R-square	Without MV			With MV
CP	0.506			0.506
EA	0.432			0.433
TP	0.499			0.500
Change in Adjusted R^2 - with and without Marker Variables
Adjusted R-square	Without MV			With MV
CP	0.499			0.498
EA	0.426			0.425
TP	0.492			0.491

Finally, the full collinearity test, following Kock (2015) which used variance inflation factors (VIF) to confirm the absence of common method bias obtained VIF values below the 3.3 threshold for every item (as shown in Table 3) thus concluding that that there were no CMV issues in the model, allowing the study to proceed with additional statistical analysis.

Table 3. Full collinearity test - VIF.

	CP	DL	EA	EV	RANDOM	TL	TP
CP					1.708
DL					1.510
EA					2.000
EV					1.155
RANDOM
TL					1.415
TP					1.904

5.3. Assessment of measurement model

5.3.1. Factors loadings, construct reliability and validity

In the study, factor loading (Table A3) was used to correlate each item with its principal component, as defined by Pett et al. (2003). Given that none of the items had factor loadings less than the recommended 0.6 (Awang, 2015; Yana et al., 2015), no items were removed. Reliability was established using Cronbach’s Alpha and Composite Reliability (CR), both of which exceeded the required 0.70 threshold (Hair et al., 2021), with values ranging from 0.750 to 0.942 and 0.810 to 0.943 (Table 4), respectively.

Table 4. Construct reliability analysis.

	Cronbach’s alpha	Composite reliability (rho_a)
Adp	0.880	0.889
CP	0.868	0.868
Cog	0.846	0.846
EV	0.929	0.933
Pro	0.825	0.883
Res	0.845	0.872
SE	0.750	0.810
TL	0.942	0.943
TP	0.905	0.911
Tec	0.913	0.923

Adp: adaptability; Cog: cognitive; Pro: proactivity, Res: resilience; SE: socio-emotional; Tec: technical.

Convergent validity was next confirmed as the average variance extracted (AVE) for all constructs was above the recommended 0.50 threshold (Table 5), indicating no issue.

Table 5. Construct convergent validity (AVE).

	Average variance extracted (AVE)
Adp	0.737
CP	0.716
Cog	0.867
EV	0.825
Pro	0.659
Res	0.688
SE	0.795
TL	0.777
TP	0.680
Tec	0.695

Adp: adaptability; Cog: cognitive; Pro: proactivity; Res: resilience; SE: socio-emotional; Tec: technical.

Finally, discriminant validity was established using the Fornell and Larcker Criterion (Table A4), cross loadings (Table A5), and the Heterotrait-Monotrait (HTMT) Ratio (Table A6). These results confirmed discriminant validity had been achieved.

5.3.2. Validating higher-order constructs

In this study, DL and EA are higher-order constructs built on the lower-order constructs of technical, cognitive and socio-emotional dimensions (for DL), and proactivity, adaptability and resilience (for EA), respectively. An evaluation of factor loadings for these constructs showed all indicators exceeded the minimum acceptable factor loading value of 0.70 (Hair et al., 2021) (Table 6), hence, no items were discarded. The reliability of the higher-order constructs was established through Cronbach’s alpha and composite reliability (Table 6), both surpassing the recommended value of 0.700 (Hair et al., 2021).

Table 6. Reliability analysis, factor loading and AVE for higher-order constructs.

	Outer loadings	Cronbach’s alpha	Composite reliability (rho_a)	Average variance extracted (AVE)
Cog <- DL	0.911	0.882	0.887	0.809
SE <- DL	0.885
Tec <- DL	0.902
Adp <- EA	0.901	0.833	0.837	0.752
Pro <- EA	0.794
Res <- EA	0.902

Adp: adaptability; Cog: cognitive; Pro: proactivity; Res: resilience; SE: socio-emotional; Tec: technical.

Average Variance Extracted (AVE) (Table 6) exceeded the acceptable value of 0.500, indicating convergent validity (Henseler et al., 2015). Discriminant validity was confirmed by contrasting the latent variable correlations with the square root of AVE (Fornell & Larcker, 1981) (Table A7), and via the Heterotrait-Monotrait (HTMT) Ratio (Table A8). These analyses affirmed the discriminant validity of the higher-order constructs of DL and EA.

5.4. Assessment of the structural model

5.4.1. Multicollinearity analysis

In the present study, Variance Inflation Factor (VIF) was used to investigate potential issues of multicollinearity. Based on recommendations by Kim (2019) and Hair et al. (2021), a VIF value exceeding 5.0 signifies the presence of multicollinearity. As depicted in Table A9, the VIF values in this study ranged from 1.434 to 4.853. These values, being under the specified threshold of 5, suggest the absence of substantial multicollinearity problems within the study.

5.4.2. Results of hypotheses testing

The results of hypotheses testing in this study are shown in Table 7 and Figure 2 below. As revealed by the statistical significance of these relationships, all three EV, DL and TL have significant and positive relationships on EA thus supporting hypotheses H1, H2 and H3. Likewise, the results also proved that EA does have a significant positive relationship on the two dimensions of JP thus further supporting H4a and H4b, respectively.

Figure 2. Structural model with interaction effect.

Table 7. Summary of total effects, direct effects and specific indirect effects of constructs.

	Total effect		Direct effect		Specific Indirect effect
	Coefficient	p-value	Coefficient	p-value		Coefficient	SD	t-value	p-value	CI (2.5%–97.5%)
DL -> CP	0.336	0.000	0.213	0.001	DL -> EA -> CP	0.122	0.045	2.704	0.007	0.046-0.216
DL -> EA	0.385	0.000	0.385	0.000	DL -> EA -> TP	0.144	0.051	2.808	0.005	0.054-0.248
DL -> TP	0.309	0.000	0.166	0.017	EV -> EA -> CP	0.093	0.027	3.418	0.001	0.048-0.153
EA -> CP	0.318	0.000	0.318	0.000	EV -> EA -> TP	0.109	0.031	3.479	0.001	0.058-0.179
EA -> TP	0.373	0.000	0.373	0.000	TL -> EA -> CP	0.058	0.026	2.280	0.023	0.018-0.119
EV -> CP	0.349	0.000	0.256	0.000	TL -> EA -> TP	0.069	0.029	2.375	0.018	0.022-0.134
EV -> EA	0.293	0.000	0.293	0.000
EV -> TP	0.344	0.000	0.235	0.000
TL -> CP	0.195	0.001	0.137	0.016
TL -> EA	0.184	0.002	0.184	0.002
TL -> TP	0.198	0.001	0.129	0.017

Next, as shown in Table 7, it was proven that there were significant partial mediating roles of EA in the respective relationships between EV, DL and TL on both TP and CP. The total effect of EV, DL and TL on TP and CP were found to be significant. Moreover, the direct effects have remained significant even with the inclusion of the mediator. Hence, it can be concluded that EA partially mediates the relationship between EV, DL and TL on TP and CP thus supporting H5a H5b, H6a, H6b, H7a and H7b, respectively.

5.4.3. Coefficient of determination (r²)

In this research, the coefficient of determination (r² value) was employed to measure the predictive accuracy of the model as recommended by Chicco et al. (2021), Sharma et al. (2020), and Zhang et al. (2018). As per Hair et al. (2021), r² value of 1 signifies perfect prediction, while 0.75, 0.50, and 0.25 can be considered substantial, moderate, and weak predictive accuracy.

As shown in Table 8, EA exhibited an r² value of 0.432 in the study, suggesting that EV, DL, and TL account for 43% of its variance. TP and CP demonstrated r² values of 0.499 and 0.506 respectively, denoting that EA explains 50% of their variance. These r² values surpass the 0.10 threshold proposed by Falk and Miller (1992), confirming that the model possesses acceptable predictive power.

Table 8. Coefficient of determination (r²).

	R-square
CP	0.506
EA	0.432
TP	0.499

5.4.4. Effect size (f²)

The study also evaluated the predictive capability of each independent construct using its effect size (f²), which estimates the predictive relevance and change in r² when a specific variable is excluded from the model, as described by Hair et al. (2021). Following the classifications provided by Cohen (2013) and Hair et al. (2021), effect sizes are categorized into small (0.02 to 0.15), medium (0.15 to 0.35), and large (0.35 and above). This effect size assessment was incorporated in the study to account for the differential impact of variables in the model as shown in Table 9.

Table 9. Assessment of effect sizes (f²).

	Employee agility (endogenous variable)		Task performance (endogenous variable)		Contextual performance (endogenous variable)
Exogenous variables	f^2 value	Effect size	f^2 value	Effect size	f^2 value	Effect size
Employee vitality	0.118	Small	0.077	Small	0.093	Small
Digital literacy	0.221		0.038	Small	0.064	Small
Transformational leadership	0.043	Small	0.023	Small	0.026	Small
Employee agility			0.158	Medium	0.116	Small

5.4.5. Predictive relevance of model (Q²)

The study also assessed the predictive relevance (Q²) of the structural model to determine its predictive power and accuracy (Al Mansoori et al., 2020; Newaz et al., 2020). This analysis, which predicts endogenous constructs in the reflective model, was conducted using PLS Predict in Smart PLS (Najib et al., 2021). Ali and Kashif (2020) suggest that the Q² value should surpass zero, with 0.02, 0.15, and 0.35 signifying weak, moderate, and substantial predictive relevance, respectively. In this study, as presented in Table 10, the Q² values for EA, TP, and CP were 0.408, 0.405, and 0.434 respectively, all exceeding the 0.35 threshold, indicating the model’s strong predictive relevance.

Table 10. Predictive relevance of model (Q²).

	Q²predict
CP	0.434
EA	0.408
TP	0.405

6. Discussions, conclusions, limitations and recommendations

The study unveils significant relationships, both direct and indirect, among factors such as EV, DL, TL, and EA. EA in turn is shown to positively influence both task and CP. Firstly, the research highlights a positive correlation between EV and EA. Vital employees, characterized by their will to grow, energy, enthusiasm, and robust mental health, exhibit a propensity for self-improvement and alignment with their work environment, thereby driving their agility (Dawis & Lofquist, 1984; Fernández-Abascal & Martín-Díaz, 2015; VanderLind, 2017). This is further substantiated by the work of Menon and Suresh (2020), which demonstrates how emotional intelligence and learning, along with innovation, complement vitality in fostering agility.

Secondly, the study underscores a substantial influence of DL on EA. This finding resonates with prior research, like that of Lai et al. (2021), which found IT competency to significantly affect task autonomy and consequently, EA. The importance of DL is further enhanced by the growing influence of Internet of Things (IoT) on white-collar professions, enabling a more resilient workplace (Abdussamad et al., 2022; Fernandez, 2020). Thirdly, the study establishes a positive relationship between TL and EA. Transformational leaders inspire positive changes, facilitate creative outcomes, and act as change agents by fostering knowledge management within organizations, empowering employees to effectively adapt and respond to changes (Cherry, 2022; Ghasabeh, 2021; Yi et al., 2019). Additionally, the study demonstrates that agility can bolster TP by improving adaptability to workplace changes, enhancing problem-solving capabilities, and promoting creativity and innovation. Finally, the indirect relationships between these factors and TP as well as CP were also identified. The study posits that EA plays a significant mediating role in these relationships.

Besides these direct relationships, this study also explores the interplay between EV, DL, TL, and JP, emphasizing the key mediating role of EA. The result demonstrates a connection between EV and both task and CP, underlining the critical mediation of EA. High vitality in HRP (characterized by energy, enthusiasm, and resilience) leads to an increased willingness to learn and adapt, fostering a positive mindset that can enhance agility. This improved agility subsequently contributes to superior TP and CP by equipping employees with better stress management abilities, improved cognitive functioning, and increased emotional intelligence as discussed earlier.

Next, the study establishes a significant relationship between DL and TP, again pointing to the mediating role of EA. DL is essential for HRP to successfully adjust to their job, leading to heightened TP. Higher DL levels also contribute to improved CP, as they promote EA, making HRP more adaptable and flexible and how this is consistent with earlier finding from like studies (Fischer et al., 2022; Lim et al., 2021; 2022; Nadzim & Halim, 2022; Saputra, 2022; Saputra et al., 2022). This aligns with the WAT and DCT, emphasizing the importance of resource development, adaptability enhancement, and job adjustment.

Lastly, the study reaffirms the mediating role of EA in the relationship between TL and TP and between TL and CP. In this respect, TL were able to increase EA leading to the ability to adapt quickly to changes and contribute to superior organizational performance, facilitating more efficient and effective work thus improving TP. Additionally, the positive environment (Kohan et al., 2018) created by transformational leaders also enhances EA among HRP (Desai, 2021), leading to higher levels of CP.

This research’s principal contribution lies in the fusion of the WAT and DCT to explore the relationships among HRP’s EV, DL, TL, and TP and CP, with a key focus on EA as a mediator. This approach broadens our understanding of factors influencing HR performance in the digital era and expands the HR literature by highlighting the essential roles of DL, leadership style, employee well-being, and agility in HRP’s JP. By introducing EV as a significant predictor of TP and CP, this study extends the WAT beyond its traditional focus on aligning individual skills, needs, and values with job requirements. It similarly enhances the DCT by investigating DL and TL’s role in HRP’s performance, extending the theory’s emphasis on organizational adaptability, learning, and innovation to individual HRP’s capabilities. Moreover, this research deepens the understanding of the less explored construct of EA by exploring its mediating role in the relationships between EV, DL, TL, and JP. These insights enrich current knowledge of the mechanisms that potentially impact HRP’s JP.

From a practical standpoint, this study offers valuable insights for HRP and the manufacturing sector by highlighting the importance of EV, DL, TL, and EA. For HR practitioners, it emphasizes the significance of maintaining energy and enthusiasm through well-being programs, work-life balance, and stress management initiatives. DL is underscored as crucial in the digital age, urging organizations to invest in training programs for HRP to enhance their skills in data analytics, HR information systems, and online collaboration tools. This proficiency in DL can contribute to better TP and CP by streamlining processes and improving communication.

In the context of rapid technological advancements, the study emphasizes the pivotal role of EA, urging manufacturing companies to ensure their HR teams possess the necessary DL and continuously improve their digital capabilities. TL is recognized as a critical factor influencing TP and CP, with leaders encouraged to adopt a transformational style to foster a positive work environment, enhance morale, and promote HR excellence. Lastly, the study identifies EA as a significant mediator between various factors, providing guidance on resource allocation for maximum impact. Manufacturing companies are recommended to prioritize investments in DL training and EA initiatives, recognizing that agile HRP can respond effectively to changing challenges and contribute to better TP and CP. This involves encouraging continuous learning, fostering innovation, and implementing flexible work policies to empower HRP to cultivate agility.

While this study presents significant theoretical, methodological, and practical contributions, it recognizes limitations, primarily with the reliance on self-reported data to gauge respondent’s beliefs and perceptions about research variables. Although employing measures for addressing potential common method bias, it suggests future studies to corroborate self-reported data with third-party performance assessments. The study, being cross-sectional, cannot establish cause-effect relationships or temporal links between variables, nor account for developmental changes and individual differences. Future research could benefit from a longitudinal design, providing more substantial evidence for causal relationships.

The study’s focus on HRP in the manufacturing sector limits its applicability, urging future research to replicate the model across various sectors for wider generalizability. Future studies could also extend the understanding of the effects of EV, DL, TL, and EA on other aspects of HRP’s professional experiences such as job satisfaction, turnover intention, job motivation, and overall well-being.

Further examination of EA’s mediating role in the context of different leadership styles and their influence on HR performance is recommended. Potential effects of these variables on additional dimensions of JP, like adaptive performance and counterproductive behavior, could be explored. Additional potential mediators such as work-life balance and employee well-being, as well as moderating factors like technological infrastructure and organizational culture, could also be considered. Given the limitations of the study’s cross-sectional design, future research should employ longitudinal studies to better identify potential causal relationships between variables.

As put forth at the beginning of this article, the JP of HRPs have proven critical to organizations especially during times of turbulence such as COVID-19. Amid the rest of the benefits and rationales of this study, perhaps the most crucial is to prepare organizations for the day when the business world encounters another major disruption ala COVID-19 sometime in the future. And this time HRPs will be more prepared.

Disclosure statement

No potential conflict of interest was reported by the author.

Data availability statement

The data that support the findings of this study are openly available in Zenodo at https://doi.org/10.5281/zenodo.8082042

Additional information

Funding

The authors received no direct funding for this research.

Notes on contributors

Ying Keat Chong

Ying Keat Chong is a graduate of the Institute of Chartered Secretaries and Administrators (ICSA). He also holds a Master of Management from the Open University of Malaysia (OUM). He’s currently pursuing his PhD in Human Resources at the School of Management, Universiti Sains Malaysia (USM) in Penang, Malaysia. Ying Keat has been the Chief Financial Officer cum National Staff Director of Dalat International School in Penang since 2008 where he is actively involved in strategic level of the company in areas of corporate finance and human resources. His research interest areas include human resources management, organisational behaviour, and strategic management.

Siti Rohaida Mohamed Zainal

Siti Rohaida Mohamed Zainal, Ph.D. is an Associate Prof. at School of Management. She graduated with a BBA from the University of Missouri Kansas City, USA and MBA degree from the Universiti Sains Malaysia (USM). She joined USM in July 2007 after completing her PhD at the Universiti Teknologi Mara, Shah Alam. Siti Rohaida is attached to Strategy and Organisational Management at USM and teaches Strategic Human Resource Management, HR Analytics, Leadership Dynamic and Strategic Leadership. Currently she is Deputy Dean of Academic, Career and International. In addition of assisting the administrative related work, she is still active with research and publication activities.

Appendices A

Table A1. Respondent profile.

	Frequency	Percentage
Gender
Male	70	23.3%
Female	230	76.7%
Age
25 and below	23	7.7%
26 to 41	153	51.0%
42 to 57	113	37.7%
58 and above	11	3.7%
Job level
Senior Manager and above	61	20.3%
Manager	75	25.0%
Senior executive	70	23.3%
Junior executive	66	22.0%
Clerical/administrative or equivalent	28	9.3%
Sector
Electrical & electronics	62	20.7%
Precision engineering and equipment	62	20.7%
Medical devices	11	3.7%
Aerospace/avionics	10	3.3%
Others	155	51.7%

Table A2. Harman’s single-factor test.

Total Variance Explained
Component	Initial Eigenvalues			Extraction Sums of Squared Loadings
	Total	% of Variance	Cumulative %	Total	% of Variance	Cumulative %
1	16.920	38.454	38.454	16.920	38.454	38.454
2	4.322	9.823	48.277
3	3.233	7.348	55.625
4	2.375	5.398	61.023
5	2.029	4.611	65.635
6	1.472	3.345	68.979
7	1.012	2.299	71.278
8	.970	2.206	73.484
9	.932	2.118	75.602
10	.777	1.767	77.369
11	.677	1.538	78.906
12	.604	1.373	80.280
13	.584	1.328	81.608
14	.547	1.243	82.851
15	.515	1.170	84.020
16	.480	1.091	85.111
17	.452	1.027	86.138
18	.441	1.003	87.141
19	.408	.928	88.070
20	.384	.873	88.942
21	.357	.811	89.753
22	.348	.790	90.543
23	.328	.746	91.290
24	.324	.737	92.027
25	.296	.674	92.700
26	.284	.645	93.346
27	.264	.599	93.945
28	.252	.573	94.518
29	.239	.542	95.060
30	.223	.507	95.567
31	.222	.505	96.072
32	.211	.479	96.550
33	.205	.467	97.017
34	.180	.409	97.427
35	.161	.365	97.792
36	.156	.355	98.147
37	.150	.341	98.488
38	.124	.281	98.769
39	.114	.260	99.029
40	.104	.237	99.266
41	.089	.203	99.469
42	.086	.196	99.666
43	.078	.178	99.843
44	.069	.157	100.000

Extraction Method: Principal Component Analysis.

Table A3. Factor loading.

	Adp	CP	Cog	EV	Pro	Res	SE	TL	TP	Tec
Adp45	0.766
Adp46	0.896
Adp47	0.923
Adp48	0.842
CP15		0.854
CP16		0.865
CP17		0.856
CP18		0.809
Cog30			0.933
Cog31			0.929
EV19				0.919
EV20				0.922
EV21				0.865
EV23				0.925
Pro41					0.901
Pro42					0.926
Pro43					0.757
Pro44					0.627
Res49						0.896
Res50						0.899
Res51						0.846
Res52						0.654
SE32							0.929
SE33							0.854
TL34								0.829
TL36								0.888
TL37								0.897
TL38								0.866
TL39								0.907
TL40								0.901
TP9									0.847
TP10									0.863
TP11									0.798
TP12									0.857
TP13									0.730
TP14									0.846
Tec24										0.793
Tec25										0.814
Tec26										0.843
Tec27										0.847
Tec28										0.844
Tec29										0.858

Table A4. Fornell & larcker criterion.

	Adp	CP	Cog	EV	Pro	Res	SE	TL	TP	Tec
Adp	0.859
CP	0.541	0.846
Cog	0.543	0.411	0.931
EV	0.395	0.531	0.198	0.908
Pro	0.538	0.525	0.264	0.455	0.812
Res	0.781	0.542	0.422	0.411	0.553	0.830
SE	0.470	0.446	0.733	0.249	0.319	0.367	0.892
TL	0.401	0.479	0.315	0.453	0.416	0.382	0.359	0.882
TP	0.567	0.766	0.408	0.520	0.525	0.558	0.411	0.469	0.825
Tec	0.549	0.501	0.736	0.297	0.326	0.452	0.672	0.335	0.469	0.834

Note: the square root of AVE is in bold and italicized.

Adp; adaptability, Cog; cognitive, Pro; proactivity, Res; resilience, SE; socio-emotional, Tec; technical.

Table A5. Cross loading.

	Adp	CP	Cog	EV	Pro	Res	SE	TL	TP	Tec
Adp45	0.766	0.403	0.330	0.386	0.416	0.587	0.348	0.330	0.438	0.391
Adp46	0.896	0.455	0.557	0.332	0.471	0.664	0.464	0.356	0.501	0.556
Adp47	0.923	0.520	0.495	0.358	0.545	0.726	0.423	0.364	0.535	0.497
Adp48	0.842	0.475	0.466	0.289	0.407	0.701	0.372	0.326	0.470	0.429
CP15	0.527	0.854	0.357	0.458	0.434	0.529	0.352	0.364	0.730	0.425
CP16	0.458	0.865	0.351	0.384	0.414	0.459	0.379	0.377	0.632	0.412
CP17	0.435	0.856	0.326	0.489	0.432	0.459	0.347	0.439	0.608	0.438
CP18	0.411	0.809	0.356	0.461	0.495	0.387	0.431	0.438	0.621	0.420
Cog30	0.514	0.386	0.933	0.213	0.249	0.400	0.671	0.314	0.381	0.679
Cog31	0.497	0.379	0.929	0.155	0.242	0.386	0.693	0.271	0.378	0.692
EV19	0.351	0.476	0.148	0.919	0.409	0.360	0.217	0.362	0.480	0.251
EV20	0.362	0.487	0.161	0.922	0.441	0.367	0.230	0.402	0.449	0.276
EV21	0.304	0.432	0.162	0.865	0.381	0.338	0.210	0.460	0.453	0.237
EV23	0.411	0.529	0.241	0.925	0.419	0.421	0.246	0.427	0.505	0.311
Pro41	0.491	0.491	0.217	0.470	0.901	0.505	0.311	0.431	0.490	0.319
Pro42	0.530	0.513	0.290	0.444	0.926	0.510	0.331	0.415	0.499	0.332
Pro43	0.358	0.367	0.172	0.277	0.757	0.384	0.182	0.270	0.370	0.242
Pro44	0.331	0.288	0.156	0.218	0.627	0.382	0.168	0.159	0.309	0.109
Res49	0.755	0.509	0.420	0.337	0.486	0.896	0.384	0.307	0.547	0.437
Res50	0.729	0.489	0.378	0.386	0.473	0.899	0.339	0.363	0.510	0.413
Res51	0.604	0.423	0.334	0.364	0.399	0.846	0.283	0.320	0.444	0.367
Res52	0.462	0.361	0.245	0.270	0.500	0.654	0.181	0.279	0.316	0.256
SE32	0.505	0.410	0.765	0.218	0.326	0.403	0.929	0.366	0.412	0.693
SE33	0.305	0.387	0.506	0.232	0.230	0.227	0.854	0.260	0.308	0.476
TL34	0.382	0.421	0.272	0.439	0.350	0.342	0.319	0.829	0.480	0.351
TL36	0.337	0.413	0.232	0.392	0.381	0.327	0.288	0.888	0.435	0.272
TL37	0.331	0.432	0.307	0.378	0.389	0.296	0.303	0.897	0.430	0.288
TL38	0.343	0.420	0.292	0.389	0.323	0.321	0.376	0.866	0.364	0.278
TL39	0.366	0.455	0.275	0.435	0.393	0.376	0.315	0.907	0.415	0.299
TL40	0.355	0.387	0.287	0.355	0.359	0.357	0.299	0.901	0.344	0.276
TP9	0.525	0.638	0.372	0.470	0.444	0.507	0.386	0.464	0.847	0.446
TP10	0.510	0.660	0.357	0.415	0.470	0.531	0.349	0.382	0.863	0.435
TP11	0.430	0.573	0.288	0.370	0.384	0.433	0.295	0.354	0.798	0.340
TP12	0.488	0.613	0.351	0.353	0.441	0.451	0.316	0.343	0.857	0.357
TP13	0.371	0.564	0.309	0.406	0.379	0.327	0.301	0.377	0.730	0.342
TP14	0.466	0.727	0.334	0.538	0.468	0.486	0.371	0.391	0.846	0.385
Tec24	0.404	0.308	0.570	0.185	0.192	0.307	0.496	0.196	0.317	0.793
Tec25	0.542	0.490	0.606	0.276	0.319	0.456	0.543	0.315	0.481	0.814
Tec26	0.489	0.519	0.551	0.348	0.354	0.424	0.541	0.329	0.471	0.843
Tec27	0.421	0.415	0.618	0.256	0.226	0.344	0.560	0.242	0.339	0.847
Tec28	0.423	0.354	0.663	0.168	0.242	0.339	0.612	0.283	0.323	0.844
Tec29	0.421	0.352	0.691	0.205	0.247	0.341	0.613	0.273	0.353	0.858

Adp; adaptability, Cog; cognitive, Pro; proactivity, Res; resilience, SE; socio-emotional, Tec; technical.

Table A6. Heterotrait-Monotrait (HTMT) ratio.

	Adp	CP	Cog	EV	Pro	Res	SE	TL	TP	Tec
Adp
CP	0.618
Cog	0.625	0.479
EV	0.438	0.588	0.221
Pro	0.619	0.604	0.308	0.496
Res	0.895	0.630	0.493	0.462	0.669
SE	0.556	0.552	0.893	0.301	0.381	0.432
TL	0.440	0.528	0.352	0.484	0.445	0.430	0.417
TP	0.632	0.861	0.465	0.562	0.594	0.624	0.485	0.504
Tec	0.599	0.546	0.841	0.310	0.346	0.496	0.791	0.352	0.498

Adp: adaptability; Cog: cognitive; Pro: proactivity; Res: resilience; SE: socio-emotional; Tec: technical.

Table A7. Fornell & larcker criterion for higher-order constructs.

	CP	DL	EA	EV	TL	TP
CP	0.846
DL	0.506	0.899
EA	0.619	0.535	0.867
EV	0.531	0.278	0.483	0.908
TL	0.479	0.374	0.460	0.454	0.882
TP	0.766	0.479	0.635	0.520	0.470	0.825

Note. The square root of AVE is in bold and italicized.

Table A8. Heterotrait-Monotrait (HTMT) ratio for higher-order constructs.

	CP	DL	EA	EV	TL	TP
CP
DL	0.575
EA	0.728	0.618
EV	0.588	0.303	0.550
TL	0.528	0.410	0.520	0.484
TP	0.861	0.531	0.728	0.562	0.504

Table A9. Multicollinearity statistics (VIF) for indicators.

	VIF
Adp45	1.648
Adp46	3.421
Adp47	3.978
Adp48	2.096
CP15	2.300
CP16	2.430
CP17	2.147
CP18	1.784
Cog30	2.162
Cog31	2.162
EV19	4.688
EV20	4.732
EV21	2.735
EV23	3.836
Pro41	4.451
Pro42	4.853
Pro43	1.685
Pro44	1.434
Res49	3.597
Res50	3.667
Res51	2.123
Res52	1.428
SE32	1.561
SE33	1.561
TL34	2.509
TL36	3.519
TL37	3.759
TL38	3.110
TL39	4.438
TL40	4.565
TP9	2.457
TP10	2.720
TP11	2.213
TP12	2.821
TP13	1.754
TP14	2.475
Tec24	2.211
Tec25	2.426
Tec26	2.968
Tec27	2.923
Tec28	4.280
Tec29	4.784

Adp: adaptability; Cog: cognitive; Pro: proactivity; Res: resilience; SE: socio-emotional; Tec: technical.