According to reviews from Amabile and Pillemer (2012), the main research orientation of early organizational creativity is to discuss creative people’s personality traits or the ability to solve problems with creativity from the aspects of trait or cognitive perspective. Later, social psychologists found from creative people’s autobiographies and letters that creative people are more inclined to produce novel and useful ideas in certain social situations. Thus, scholars’ assessments for individual creativity also gradually shift from the emphasis on individual cognitive competence to the impact of social situational factors on individual creativity performance or IB (Orfila-Sintes and Mattsson, 2009; Chen and Zhou, 2017). Creativity is interpreted from the view of behaviors, which refers to employees who come up with novel or useful ideas, while IB includes the process in which employees propose, introduce, or utilize new ideas in the workplace, and implement creative ideas in different ways with subsequent purposes (Reade and Lee, 2016; Chen and Zhou, 2017). Creativity means that an individual puts forward novel or useful ideas, problem-solving methods, or processes (Amabile, 2011). However, IB refers to employees who propose, introduce, or apply new ideas at work, which will be further implemented or fulfilled through different ways (Yuan and Woodman, 2010; Reade and Lee, 2016). The difference between IB and creativity of employees lies in the fact that IB focuses on the occurrence and implementation of employees’ new ideas. In other words, IB includes creative thinking and concept practice, so creative power or creativity can also be regarded as one of the types of IB (Yuan and Woodman, 2010; Kao et al., 2015). Regarding it from the depth of creativity, it can be divided into big creativity, which changes human life and civilization, and small creativity, which improves the quality of individual work or life and solves daily problems, also known as daily creativity (Conner and Silvia, 2015). Despite members in different professional fields showing unique behaviors of problem discovery and problem solving due to the characteristics of their work situations (Kaufman and Baer, 2005), these behaviors still have common characteristics. For example, employees in daily work ponder something out of framework or reorganize existing ideas, seek or apply new technologies, new procedures, and new approaches at work, figure out creative ideas, sell new ideas to others, and actively strive for resources needed to fulfill new ideas, and plan a timetable to accomplish new ideas (Scott and Bruce, 1994).

For a long time, employees in work surroundings have been struggling with physical and mental stress which keeps employees from coping with learning challenges in a positive manner (Ahmed and Nawaz, 2015). Bewick et al. (2010) argued that British employees were taken as the research object in a study, compared with their peers, they often have considerable pressure on loans, life, and performance, and it was emphasized that scholars are necessarily supposed to shift the focus from work performance to the exploration of employees’ psychological issues (Kurtessis et al., 2017; Meyers et al., 2019). Even though scholars have explored employees’ IB from different levels, some research gaps still exist which are worthy of being explored and discussed, such as how IB develop, and internal and external elements influencing employees’ IB (Reade and Lee, 2016; Chen and Zhou, 2017; Baradarani and Kilic, 2018). Besides, Folkman and Moskowitz (2000) indicated in their study that the subsequent research needs to emphasize the discussion of positive emotions and IB (Kurtessis et al., 2017) as figuring out relevant elements available to keep down psychological health problems arising from stress in as an effective way as possible, if it is not explored from the aspect of positive results (Thompson et al., 2016). Thus, based on the SCCT, the study adopts IB as the outcome variable for the exploration of the effect that correlative factors bring to it. Figuring out different mechanisms conducive to employees’ IB is in the interest of organizational behaviors and administrators.

For the past few years, scholars have been more committed to conducting research related to employability (Ineson et al., 2013; Thompson et al., 2016). The concepts and operations of industrial organizations worldwide have been modified by the substantial technological, social, and economic vicissitudes that have sprung up in recent decades (Abbas et al., 2015; Akgunduz et al., 2018; Abbas and Sağsan, 2019). Therefore, the highest standards of human capital development are guaranteed by dynamic organizations, which make contributions to economic progress (Ahmed et al., 2015; Baek and Cho, 2018). Scholars have conducted a study on the implications of EE and the causal relationships between EE and other factors (Hennemann and Liefner, 2010; Thompson et al., 2016; Baek and Cho, 2018) by the ways of research situations and method design, as well as the integration of theoretical and practical analysis (Ineson et al., 2013). Van Der Heijde and Van Der Heijden (2006) stated that EE is the proper application of individual capabilities (Pan and Lee, 2011; Blázquez et al., 2018), constant acquisition and creation of necessary occupational skills to fulfill all the tasks, and adapt to internal and external changes in the job market (Chang and Edwards, 2015; Akgunduz et al., 2018). Thereby, the demand for a critical and reflective mind, capabilities of solving problems, self-government, learning, and related capabilities are constantly enhanced in an interdisciplinary way (Ineson et al., 2013; Thompson et al., 2016; Makkonen and Olkkonen, 2017). Some previous studies have stated that, besides the influence brought to EE by basic education, elements such as individual conditions, interpersonal relationships, and external elements that are not accessible in human resources need to be taken into consideration as well. Pan and Lee (2011) conducted a survey of the samples in Taiwan and adopted the scale of employment from Andrews and Higson (2008), who suggested that employability necessarily involves the general and professional capabilities required at work, attitude to work, occupational plan capabilities, and confidence. The classification of employability made by Pan and Lee (2011) is taken as the measure for EE in this study.

According to De Cuyper et al. (2008), EE is of great importance in the society of post-industrial knowledge, which constantly updates knowledge to keep competitive in a worldwide market and makes them accessible to handling temporary and subsequent development—new psychological contracts developed by individuals tend to enhance their IB (Lent et al., 2011; Ahmed and Nawaz, 2015; Akgunduz et al., 2018). Besides, with less time, related experience, skills, and knowledge that have been updated, individuals are available to process the same things and tasks in a more effective way (Lent et al., 2011; Chang and Edwards, 2015) —and a social network that has undergone positive development—to increase EE. Abundant time saved will be contributed to life needs and individual planning for future, thus strengthening IB (Thompson et al., 2016). Likewise, higher employability can enable employees to contend with job challenges in the future with a broader view. They not only master the content of organizational tasks, but also show a more precise direction for planning and preparing for tasks to be accomplished (Ineson et al., 2013; Chang and Edwards, 2015), thus keeping down their insecurity and improving IB. Based on the above phenomena, the study proposes the following hypothesis:

According to SCCT scholars, both environmental factors and cognitive factors in a certain context, particularly those beliefs leading to success and behavior, will influence individuals’ behavioral outcomes (Brown et al., 2011; Chin and Rasdi, 2014; Chang and Edwards, 2015; Liguori et al., 2019). These beliefs are called “self-efficacy” by them, namely a significant cognitive variable in individual factors during accounting for individual behaviors (Caesens and Stinglhamber, 2014), and interaction with the surroundings (Lent et al., 2011; Duffy et al., 2014; Chang and Edwards, 2015; Jemini-Gashi et al., 2019). It can also be regarded as the foundation for the motivation of human behaviors (Cordova et al., 2014), mental health, and individual accomplishments (Lent et al., 2011; Liguori et al., 2019). The field of human resources takes a wide application of self-efficacy to probe into the psychological cognitive factors of employees in different situations and their positive impact on task accomplishment and employees’ occupational development (Brown et al., 2011; Caesens and Stinglhamber, 2014; Duffy et al., 2014; Jemini-Gashi et al., 2019). To have a clearer understanding of the application of SCCT, major findings of relevant studies are described and summarized as follows in Table 1.

Based on the above discussion, it is considered that employees having confidence in their capabilities will lead to behaviors that are more efficient and interpersonal relationships that are better than whose who lack confidence (Brown et al., 2011; Chin and Rasdi, 2014; Chang and Edwards, 2015). As Chin and Rasdi (2014) considered, employees who are highly self-motivated seek resources and opportunities to fulfill tasks existing in a social network (Lent et al., 2011; Thompson et al., 2016). Only by establishing and insisting on network relationships can they achieve their goals. Knowledge and resources are in need (Lent et al., 2011; Jemini-Gashi et al., 2019). Moreover, teamwork can also be viewed as a strong network relationship, and the process in which problem solving and task fulfillment are conducted for employees via teamwork will play a positive effect on their EE (Duffy et al., 2014; Chang and Edwards, 2015). Given the above, this study puts forward H2 as follows:

Some scholars have attached their research to the concerns for psychological health, POS (Chin and Rasdi, 2014), and lifestyles for employees (Lent et al., 2011). Nonetheless, few studies yet have addressed general self-efficacy and IB in this population. Research results from Jemini-Gashi et al. (2019) showed that individuals express a lower support level, limited sources that support comes from, and low perceived support (Brown et al., 2011). According to Caesens and Stinglhamber (2014), employees who have a high level of self-efficacy are inclined to gain diversified benefits at work which eventually give rise to a higher level of work satisfaction. It indicates that employees’ failure to receive timely and necessary mental support when encountering work pressure leads to the deduction in employees’ general self-efficacy and IB (Thompson et al., 2016). Besides, it might be conducive to unique stressors. On the contrary, employees owning higher self-efficacy show a higher level of IB. In a word, the study deduces H4:

Two causal mechanisms are conducive to the development of IB in human resources, and they are PK, as well as POS. With support for IB establishment, organizations or supervisors have access to devising the organizational context, including individual and organizational factors (Chin and Rasdi, 2014; Chang and Edwards, 2015; Thompson et al., 2016; Liguori et al., 2019) to improve the efficiency and responsiveness of knowledge acquisition. According to scholars, organizations or supervisors claimed to make use of, integrate, and rearrange individual and organizational elements to establish an optimal organizational environment for building employees’ IB (Lent et al., 2011; Chin and Rasdi, 2014; Ahmed and Nawaz, 2015; Kurtessis et al., 2017; Akgunduz et al., 2018). Organizations or supervisors carry out a range of support activities to identify individual and organizational elements (Thompson et al., 2016; Liguori et al., 2019), where PK focuses on perceiving knowledge and skills that internally exist (Ineson et al., 2013; Williams and Lombrozo, 2013; Cordova et al., 2014; Li et al., 2015; Hajizadeh and Zali, 2016) and POS focuses on offering tangible and intangible resources to accelerate employees’ capabilities to fulfill their tasks or goals (Caesens and Stinglhamber, 2014; Ahmed and Nawaz, 2015; Lamm et al., 2015; Liguori et al., 2019). In this study, a better way to construct IB to make support activities adapted to PK and POS is considered.

The purpose of this research is to explore the employee IB, and to analyze the impact of internal and external factors provided by the organizations and individual knowledge base. The research sample in this study comprised employees. Purposive sampling was adopted. However, this sampling suffers from several disadvantages. Vulnerability to errors in judgment by researchers, low level of reliability and high level of bias, and inability to generalize research findings are three main disadvantages. To avoid these disadvantages, some conditions were set during sampling in this study to make the samples obtained better conform to sample reliability and to improve the generalization of the study. The results indicated that subject did not significantly impact the research variables, and so did not need to be included as an independent variable in subsequent analyses. To discuss IB of employees in a more clarified manner, not all the employees are regarded as the study population, but only the employees in the information service industry. Moreover, while filling the questionnaire, all the samples were already at work, rather than being isolated at home. In the questionnaire, participants were informed of the research purpose, research ethics, and low risks, and the questionnaire information was processed in an anonymous way. The response period ran from May 2020 to August 2020. Since different pandemic prevention policies were adopted in Taiwan, mainland China, and Malaysia, we try to explore whether these pandemic prevention policies have had a different impact on attitudes and behaviors of employees during this period. This study constructed a structural model to explore the correlations among POS, PK, self-efficacy, EE, IB, and JP. It sampled from Taiwanese, Malaysian, and mainland China companies. This study selected more than 20 Taiwanese, Malaysian, and mainland China companies, and then sent 1000 questionnaires to each of them. Finally, a total of 640 Taiwanese questionnaires, 450 Malaysian questionnaires, and 568 mainland China questionnaires were returned, for an effective response rate of 64.0, 45.0, and 56.8%. In the Taiwanese sample, most are men (61.8%), whose level of education is mostly undergraduate or above (78.9%), and most of them are between 30 and 40 years old (77.3%) with an average working year of 3.9. In the Malaysian sample, most are men (55.4%), whose level of education is mostly undergraduate or above (68.4%), and most of them are between 35 and 40 years old (43.2%) with an average working year of 5.2. In the sample of mainland China, most are men (62.1%), whose level of education is mostly undergraduate or above (66.9%), and most of them are between 30 and 35 years old (53.8%) with an average working year of 4.3.

This study hid the names of constructs and assigned the question items randomly to prevent common method variance (CMV). The Harman one-factor analysis method as used to test for CMV. The explained variance in one factor was 35.27%, which is smaller than the recommended threshold of 50%. Therefore, CMV was not problematic in this study (Podsakoff and Organ, 1986). Before conducting hypotheses testing, this study must ensure that the values of the variance inflation factor (VIF) are less than 5, but the research results showed that the VIF values were between 1.332 and 2.723. Thus, there were no multicollinearity problems among the latent variables (Hair et al., 2017).

Most of the scales in the questionnaire are adopting previous studies and modified to suit the research context. In PK, 10 items were developed based on a prior scale proposed by Silva et al. (2013). To divide POS into supervisor and colleague support (four items) and organizational support (eight items), we adopted the scale proposed by De Vos et al. (2011). In employee employability, the scales proposed by Pan and Lee (2011) were adopted, including general ability for work (GAW) (eight items), professional ability for work (PAW) (four items), attitude at work (AW) (three items), and career planning and confidence (CPC) (three items). For self-efficacy, the scale is revised and integrated with six items developed by Rigotti et al. (2008). IB was measured using Kao et al.’s (2015) instrument, which comprehensively assesses IB in three items. For JP, five items were selected based on Janssen’s (2001) scale. All items were measured with a 5-point Likert scale (1 = totally disagree; 5 = totally agree) and are shown in Table 2.

All scales used in this study were found to be reliable, with Cronbach’s α ranging from 0.83 to 0.96. Table 3 shows the reliability of each scale, and the factor loadings for each item therein. To gauge validity, this study employed confirmatory factor analysis (CFA) using AMOS 23.0 to verify the construct validity (both convergent and discriminant) of the scales. According to Hair et al.’s (2010) recommended validity criteria, CFA results show standardized factor loading of higher than 0.5; average variance extracted (AVE) ranges between 0.514 and 0.803; and composite reliability (CR) ranges between 0.863 and 0.962. All three criteria for convergent validity were met, and correlation coefficients were all less than the square root of the AVE within one dimension, suggesting that each dimension in this study had good discriminant validity.

Before proceeding to examine the structural model, we first tested model fit. Henseler et al. (2015) proposed three model fitting parameters: the standardized root mean square residual (SRMR), the normed fit index (NFI), and the exact model fit. In this study, the SRMR value was 0.054 (<0.08) and the NFI was 0.932 (>0.90) and the d_ULS < bootstrapped HI 95% of d_ULS and d_G < bootstrapped HI 95% of d_G indicating the data fits the model well. Partial least squares structural equation modeling (PLS-SEM) was adopted to construct the structural model; specifically, verification of the structural model was performed using SmartPLS 3.0 (path analysis). To assess the structural model, Hair et al. (2017) suggested looking at the R², beta (β), and the corresponding t-values via a bootstrapping procedure with a resample of 5000. They also suggested that in addition to these basic measures, researchers should also report the predictive relevance (Q²) as well as the effect sizes (f²). Prior to hypotheses testing, the values of the variance inflation factor (VIF) were determined. The VIF values were less than 5, ranging from 1 to 1.914. Thus, there were no multicollinearity problems among the predictor latent variables (Hair et al., 2017).

Figures 2–4 show the results of the hypothesized relationships and standardized coefficients in Taiwanese and mainland China samples. The results showed that IB was positively and significantly related to JP (β_Taiwan = 0.601, f² = 0.566, p < 0.001; β_China = 0.736, f² = 1.185, p < 0.001; β_Malaysia = 0.471, f² = 0.285, p < 0.001), supporting H1; respectively, we have found comparable results with Baradarani and Kilic (2018). Self-efficacy (β_Taiwan = 0.535, f² = 0.399, p < 0.001; β_China = 0.363, f² = 0.159, p < 0.001; β_Malaysia = 0.523, f² = 0.353, p < 0.001) and EE (β_Taiwan = 0.276, f² = 0.106, p < 0.001; β_China = 0.182, f² = 0.040, p < 0.001; β_Malaysia = 0.221, f² = 0.063, p < 0.001) were also positively and significantly related to IB, supporting H2 and H4. The findings of this research confirm consistent with findings of Wei et al. (2020) that self-efficacy plays a significant role in the development of IB.

In addition, self-efficacy (β_Taiwan = 0.330, f² = 0.106, p < 0.001; β_China = 0.049, f² = 0.040, p < 0.1; β_Malaysia = 0.302, f² = 0.063, p < 0.001) was positively and significantly related to EE in Taiwanese and Malaysian samples rather than in the mainland China sample, partial supporting H3. This is consistent with the results of Liu et al. (2020) and Zhao et al. (2021) that found a positive effect of self-efficacy on employability. Similarly, the paths of PK → self-efficacy (β_Taiwan = 0.271, f² = 0.104, p < 0.1; β_China = –0.052, f² = 0.003, p > 0.1; β_Malaysia = 0.302, f² = 0.110, p < 0.001) and POS → EE (β_Taiwan = 0.128, f² = 0.016, p < 0.1; β_China = 0.036, f² = 0.003, p > 0.1; β_Malaysia = 0.218, f² = 0.069, p < 0.001), showed that the relations were positive and significant in Taiwanese and Malaysian samples rather than in the mainland China sample, therefore, partially supporting H6 and H8. A similar result was found in a study of United Kingdom, Australia, and Switzerland institutions, where PK was found to have a positive influence on self-efficacy (Ineson et al., 2013).

Finally, the paths of PK → EE (β_Taiwan = 0.271, f² = 0.150, p < 0.1; β_China = –0.052, f² = 0.945, p > 0.1; β_Malaysia = 0.302, f² = 0.179, p < 0.001) and POS → self-efficacy (β_Taiwan = 0.512, f² = 0.370, p < 0.1; β_China = 0.504, f² = 0.306, p > 0.1; β_Malaysia = 0.318, f² = 0.122, p < 0.001) showed that the relations were positive and significant in both samples, supporting H5 and H7. The Stone–Geisser Q2 values obtained through the blindfolding procedures for self-efficacy (Q² = 0.178), EE (Q² = 0.335), IB (Q² = 0.352), and JP (Q² = 0.303) were larger than zero, supporting the model has predictive relevance (Hair et al., 2017).

It was confirmed that the measurement pattern was stable. However, to avoid overgeneralizing the data-driven patterns and theories, the study followed the suggestion of Hair et al. (2010) to divide the sample data into three groups based on regions (623 Taiwanese, 440 Malaysian, and 513 mainland China employees, respectively). The partial measurement invariance was established that was the basic requirement to compare as well as interpret the PLS-SEM’s findings for examining the specific Multiple Group Analysis (MGA) group’s differences (Henseler et al., 2016). Table 4 indicates the structural models’ results and MGA by using non-parametric methods including Henseler’s MGA as recommended by Henseler et al. (2009). Despite several differences in terms of significant path estimates between the groups, as indicated in Table 4, the multi-group permutation tests (right column) showed there are seven significant differences between the two groups on all paths. The results signify that the region plays a moderating role on the relationship among PK, POS, self-efficacy, SE, IB, and JP (Hair et al., 2017). The differences in paths comparison among Taiwan vs. mainland China, Taiwan vs. Malaysia, and Malaysia vs. mainland China show that four paths, two paths, and five paths were significant sequentially. These results imply that the research framework did differ among the three regions.

Based on our results, this study suggests some significant practical implications to improve the quality of human resources. First, POS and PK were equally significant and predictive for employees’ own perceived degree of self-efficacy and EE, thus having an effect on IB. Building mechanisms of mentality that are individual and organizational are conducive to employees in terms of acquiring more resources and psychological support, which provide conditions essential for IB improvement. Thereby, as countries and regions worldwide are undergoing a struggle with the COVID-19 pandemic at present, when facing such similar events, organizations need to facilitate supervisors to establish a positively close connection with employees, set up platforms for communication via technological media and information technology devices, and offer real-time tasks or psychological support.

Second, external environmental elements, the worldwide pandemic of COVID-19 in particular, may affect employees’ work state. Thus, the examination for a sense of risk management is essential for managers. Based on this, companies or organizations are suggested to turn to preventive measures for risk management in this study to contend with threats and challenges arising from adaptive risks when encountering similar events. Even though all employees are prompted to engage in online working due to this event, not all employees possess technological media or information technology devices which are required. As a result, it is a necessity for managers to keep statistics on how many employees own information technology devices first and then figure out whether work tasks can be accomplished via online working; then the work tasks that fail to be accomplished via online work need to be rearranged in accordance with a schedule.

Third, in view of the structural patterns for three regions, IB originating from self-efficacy of employees from Taiwan and Malaysia is superior to that of employees from Chinese mainland. It is found that working online will influence employees. In regions that have been blocked for a longer time, employees are likely to feel more helplessness, disability, and anxiety. Even though employees are confident in task accomplishment, they are suffering from negative energy led by blockage. In this study, managers are suggested to provide support in other ways, such as opportunities, resources, and autonomy, to assist employees in conquering the threats and challenges from their surroundings and participate in their IB.

The research findings make contributions to the literature concerning employees in specific regions, SCCT, and employees’ IB. However, there are still some limitations existing and representing subsequent research directions. First, there is considerable status for SCCT in the field of psychology, but merely a few studies have taken the relationship between building mechanism and IB of employees into consideration. Despite this study referring to the SCCT and establishing the building mechanism, and significant organizational theories are available to be drawn from the findings, other motivation theories, including the theories of organizational learning, self-efficacy, and hierarchy needs, still apply to explaining how to stimulate IB for employees in a specific region. Therefore, subsequent research is suggested to apply diversified theoretical models to identifying related psychological dimensions that play an effect on employees’ IB. Second, employees are required to do a self-report of details regarding their mental building mechanism as the indicator in the study, which is largely attributed to the actual data that is confidential and not accessible in an easy way. Nonetheless, there may be errors occurring in employees’ self-statement of mental conditions. If the actual mental conditions of employees are assessed, the connection between building mechanism and IB may be better understood, considering research ethics. In addition, subsequent researchers are suggested to include contents of interviews and employees’ observations of work state into their studies to sustain the research findings and draw a comprehensive judgment. Third, restricted by time and space, a total of 1576 valid copies of the questionnaire were sampled. The research objects were classified into employees from Taiwan and Chinese mainland. Subsequent research can be made to both expand the quantity of samples and research representativeness, and conduct an exploration and comparison of other groups, so that extra insights related to organizational behavior management are offered.