Cognitive–affective system theory

The cognitive–affective system theory was proposed by Mischel and Shoda (Reference Mischel and Shoda1995) in the 1990s, which suggests that the processing of original information into behavior encompasses two processes. The first process is encoding, in which the original information is interpreted and encoded within the cognitive–affective unit. The second process involves behavior generation, where interactions within the cognitive–affective units lead to the behavioral outcomes. This theory highlights that individuals’ reactions to situations are not passive or apathetic; instead, they are active, goal-oriented, and capable of planning and making changes on their own (Ali, Wang, & Boekhorst, Reference Ali, Wang and Boekhorst2021). These foundational assumptions align with the formation mechanism of TCB. Thus, the present paper considers individuals and their teams as rational and emotional entities and explores the underlying mechanism of TCB through the lens of cognition and emotion pathways.

SL and TCB

TCB is an important challenging organizational behavior, which is characterized by collaborative interactions among team members (Ali, Wang, & Boekhorst, Reference Ali, Wang and Boekhorst2021), the use of technologies such as social media has the potential to influence social dynamics, foster collaboration, and impact social interactions within the team (Morrison & Phelps, Reference Morrison and Phelps1999). TCB will bring new ideas and new changes to the organization, thus it has attracted more and more attention from theorists and practitioners (Zhang, Law, & Wang, Reference Zhang, Law and Wang2020; Kim & Liu, Reference Kim and Liu2017). As such, how to make this kind of behavior outside the work role as the behavior within the scope of work role is the focus of managers (Xu, Loi, & Chow, 2022; Li, Zhang, & Tian, Reference Li, Zhang and Tian2016). The existing studies on TCB of team members are still in infancy (Ngo, Mustafa, & Butt, Reference Ngo, Mustafa and Butt2022; Xu, Loi, & Chow, Reference Xu, Loi and Chow2023). The discussion on its causes mainly focuses on two aspects, individual factors and contextual factors (Xu, Loi, & Chow, Reference Xu, Loi and Chow2022; Zhou et al., Reference Zhou, Liu, Li, Cheng and Hu2020; Wang et al., Reference Wang, Wang, Li and Wang2020; Kim & Liu, Reference Kim and Liu2017; Cai et al., Reference Cai, Huo, Lan, Chen and Lam2018). For example, leadership-related factors such as benevolent leadership (Xu et al., Reference Xu, Zhao, Xi and Zhao2018), servant leadership (Xu, Loi, & Chow, Reference Xu, Loi and Chow2022), inclusive leadership (Wang et al., Reference Wang, Wang, Li and Wang2020), transformational leadership (Li et al., Reference Li, Dan, Kirkman and Xie2013), self-sacrificial leadership (Li, Zhang, & Tian, Reference Li, Zhang and Tian2016), and empowering leadership had been discussed (Zhang et al., Reference Zhang, Qian, Wang and Chen2020). However, less is known about the relationship between informal leadership and TCB. With the increasing emphasis on teammate cooperation and coordination in current organizations, the practice of informal leadership (e.g., SL) has increasingly become more prevalent (Wang & Peng, Reference Wang and Peng2022; Liang et al., Reference Liang, van Knippenberg and Gu2021).

The era of information technology and knowledge economy as the pivotal competitive advantage has made the nature of corporate work change (Cao & Ali, Reference Cao and Ali2018; Mcfarland & Ployhart, Reference Mcfarland and Ployhart2015). Within this context, organizational work has shifted to a team-based structure, especially focusing on self-managing teams (Lyndon, Pandey, & Navare, Reference Lyndon, Pandey and Navare2020). Leadership research has been moving toward a dynamic and complex adaptation process, hence the concept of SL came into being (Asa & Ajbr, Reference Asa and Ajbr2016; Carson, Tesluk, & Marrone, Reference Carson, Tesluk and Marrone2007). The advantage of SL is precisely that it breaks down the rigid identity boundaries of traditional formal leadership, inspires enthusiasm, and gives full play to the knowledge and ability of team members, which effectively makes up for the shortcomings of formal leadership (Liang et al., Reference Liang, van Knippenberg and Gu2021; Asa & Ajbr, Reference Asa and Ajbr2016). Research on the concept of SL can be summarized into three perspectives: process, capability, and structure. From the perspective of process, scholars suggested that SL is a dynamic interactive influence process in which team members lead each other to achieve team or organizational goals (Conger & Pearce, Reference Conger and Pearce2004; Derue & Ashford, Reference Derue and Ashford2010). According to Nicolaides et al. (Reference Nicolaides, Laport, Chen, Tomassetti, Weis, Zaccaro and Cortina2014), SL is a sequence of interactive influence processes in which team leadership is willingly shared among team members to achieve team goals. From the perspective of capability, scholars advocate that SL is a team capability formed by the mutual influence of team members (Pitelis & Wagner, Reference Pitelis and Wagner2019). Lastly, from the perspective of structure, scholars consider SL as a leadership structure attribute that spreads leadership across multiple team members. For instance, Carson, Tesluk, and Marrone (Reference Carson, Tesluk and Marrone2007) proposed that SL is a team emergent network attribute formed by the distribution of leadership influence among multiple team members. This study draws on the definition from the structural perspective and proposes that SL is an interactive relationship network in which each member interacts, coordinates, and switches between the roles of ‘leader and follower’ to attain team objectives. SL, as opposed to the typical centralized system in which only one person serves as the leader, allows them to transition between the positions of ‘leader and follower’ (Lee, Lee, Seo, & Choi, Reference Lee, Lee, Seo and Choi2015; Lyndon, Pandey, & Navare, Reference Lyndon, Pandey and Navare2020), thus bringing high autonomy, high ability perception, and mutual cooperation to them, which is conducive to individuals’ TCB (Chiu, Owens, & Tesluk, Reference Chiu, Owens and Tesluk2016; Zhang, Law, & Wang, Reference Zhang, Law and Wang2020).

Based on cognitive–affective system theory, leadership and other contextual factors improve or reduce team members’ performance through individual cognition and behavior (Mischel & Shoda, Reference Mischel and Shoda1995). We use it as a theoretical basis to better understand how SL, a contextual influencer, affects members to engage in TCB. TCB is characterized by actively confronting and changing the status quo with certain risks (Morrison & Phelps, Reference Morrison and Phelps1999). Participation in this behavior, therefore, depends on their perceptions of the safety of the work environment, and this belief depends largely on leadership style. Specifically, SL may motivate TCB among team members in two ways. First, SL, in contrast to traditional vertical leadership, promotes team members’ active and voluntary involvement in team management, taking up leadership responsibilities and finding solutions to problems (Chiu, Owens, & Tesluk, Reference Chiu, Owens and Tesluk2016; Nicolaides et al., Reference Nicolaides, Laport, Chen, Tomassetti, Weis, Zaccaro and Cortina2014). Along the way, those who assume the role of ‘leader’ take on the role of formal leadership, guide the team to achieve goals, and take a proactive approach to improve team effectiveness (Lee et al., Reference Lee, Lee, Seo and Choi2015). Meanwhile, members who take on the position of ‘follower’ will actively participate in team management in order to help the team develop (Derue & Ashford, Reference Derue and Ashford2010; Jonsson, Reference Jonsson2012). Second, SL focuses on mutual leadership, which promotes mutual trust among team members (Carson, Tesluk, & Marrone, Reference Carson, Tesluk and Marrone2007). Individuals have more confidence to try new things when they perceive a high degree of trust and support from other members, because others will accept and understand their mistakes (Dufour, Maoret, & Montani, Reference Dufour, Maoret and Montani2020). In addition, as members lead each other, they will also try to trust the ‘leader’ members, thereby motivating those leaders to engage more openly in TCB (Zhang, Law, & Wang, Reference Zhang, Law and Wang2020). Therefore, we propose the first hypothesis:

The mediating role of PIS

TCB is characterized by actively confronting and changing the status quo with certain risks (Morrison & Phelps, Reference Morrison and Phelps1999). Participation in this behavior, therefore, depends on their perceptions of the safety of the work environment, and this belief depends largely on leadership style. According to cognitive–affective system theory, individuals are proactive and goal-oriented, taking the initiative to plan and make changes. This aligns with the formation mechanism of TCB, which views individuals and their teams as both rational and emotional entities, and the underlying mechanism of TCB is established through the cognition path. PIS refers to employees’ belief in being insiders within a specific organization (Stamper & Masterson, Reference Stamper and Masterson2002). Within the organization, leadership will have an impact on the strength of team members’ PIS. Furthermore, leadership influences the innovative behavior of team members by improving team members’ PIS (Wang & Kim, Reference Wang and Kim2013). PIS starts from the cognitive state of team members and plays a vital role in explaining the internal mechanism of members’ cognitive process under SL (Aryee & Chen, Reference Aryee and Chen2002). SL decentralizes leadership rights and obligations to team members, so that they can establish a psychological contract with the team, thus perceiving themselves as ‘insiders’ (Asa & Ajbr, Reference Asa and Ajbr2016). When team members view themselves as ‘insiders’ of the organization through self-cognition, they will generate corresponding role identities and assume their proactive behaviors (Brown, Reference Brown2020; Mustafa, Badri, & Ramos, Reference Mustafa, Badri and Ramos2022). Team members will also continuously improve themselves and strengthen their self-efficacy, thus increasing their self-cognition (Yang & Shi, Reference Yang and Shi2020).

In addition, according to the positive distinctiveness of social identity theory, individuals tend to seek out performance in groups in order to reinforce their self-concepts (Hogg, Abrams, & Brewer, Reference Hogg, Abrams and Brewer2017). SL leads to a higher sense of insider identity (Shao, Feng & Wang, Reference Shao, Feng and Wang2016). In order to reinforce this concept of identity, they contribute more actively to the team (Hogg, Abrams, & Brewer, Reference Hogg, Abrams and Brewer2017). That is, SL leads to a higher sense of insider identity, thereby exhibiting more TCBs and forming their own personal identity position (Carson, Tesluk, & Marrone, Reference Carson, Tesluk and Marrone2007; Conger & Pearce, Reference Conger and Pearce2004). Moreover, members who regard themselves as ‘insiders’ will have a more positive emotional relationship with the team, and they are more willing to contribute time and energy to fulfill job requirements (Wang & Kim, Reference Wang and Kim2013). As such, they will consciously assume the job responsibilities, accept challenging tasks, and eventually implement more TCB. Consequently, we hypothesize:

The mediating role of EI

EI refers to an individual’s ability to regulate their own emotions and the emotions of others, as well as differentiate between different emotions. It also involves using emotional information to guide one’s thinking and actions (Mayer et al., Reference Mayer, Salovey, Caruso and Sitarenios2003; Lam & Kirby, Reference Lam and Kirby2002). Self-determination theory provides a useful framework to explain why people have different emotional states when faced with the same task activity (Deci & Ryan, Reference Deci and Ryan2000). It believes that autonomy, ability, and belonging are the three basic psychological needs of individuals, which can best satisfy the internal motivation of an individual’s work, thereby enhancing the initiative of work. Based on the above theory, we propose that, first, SL encourages individual thinking and empowers team members to thoroughly deconstruct the cognitive model, and uses a more autonomous workspace, which is conducive for team members to handle the sense of conflict (Jamshed & Majeed, Reference Jamshed and Majeed2019). Second, SL demonstrates TCB that deviates from established norms by breaking rules, constraints, and routines (Li, Zhang, & Tian, Reference Li, Zhang and Tian2016). Meanwhile, SL inspires team members to bear the responsibility, find the best solutions, and challenge the status quo (Asa & Ajbr, Reference Asa and Ajbr2016). Thereby, members enhanced their confidence and ability to work in difficult situations and acquired additional basic psychological needs ability (Hogg, Abrams, & Brewer, Reference Hogg, Abrams and Brewer2017). Under the above circumstances, the three basic psychological needs of team members are all satisfied, so that team members tend to conduct emotion cognition, emotion understanding, emotion application, and emotion management independently rather than passively, which is conducive to shaping a high level of EI (Deci & Ryan, Reference Deci and Ryan2000; Zhang et al., Reference Zhang, Qian, Wang and Chen2020).

Meanwhile, EI can effectively enhance team members’ TCB. First, EI can help team members understand their own feelings and handle them well. The tension and conflict of implementing TCB can be stressful. Thus, high EI can help team members better and faster control their emotions, muster the courage to face problems, encourage themselves to challenge obstacles, and proactively implement TCB. Second, EI can help team members better perceive others’ emotions and the surrounding environment. Team members with high EI have stronger empathy, will interact smoothly with other members in critical moments from the perspective of others and team interests, and will implement TCB that is beneficial to the team. Members with high EI do not complain about their environment and can provide help and encouragement when team members feel frustrated. Finally, EI can motivate team members to be more creative and in better shape to complete their work. This can ensure that team members are persistent and think about using more flexible solutions when encountering difficulties during the implementation of TCB and continue to adopt TCB in order to achieve the common goals of the team.

In conclusion, SL promotes the development of team members’ emotional cognition, understanding, application, and management skills, ultimately enhancing their EI (Shariq, Mukhtar, & Anwar, Reference Shariq, Mukhtar and Anwar2019). Consequently, team members gain the ability to effectively utilize emotional resources during dynamic and complex change situations. They proactively challenge the status quo, fostering a continuous inclination toward taking charge. Thus:

The moderating role of SMU

As a lightweight means of communication, team social media is not only a work platform but also a social tool for team members (Ali-Hassan, Nevo, & Wade, Reference Ali-Hassan, Nevo and Wade2015). It provides a forum for knowledge and information sharing, discussion, and co-creation (Zhou & Mou, Reference Zhou and Mou2021). SMU allows multiple members to participate in the same conversation at the same time, resulting in a network of shared understanding (Ali-Hassan, Nevo, & Wade, Reference Ali-Hassan, Nevo and Wade2015). SMU, as a prominent IT context, may instigate the effectiveness of SL in teams. Studies suggest that team members have specific goals in their choice of SMU, including seeking information and knowledge, enhancing social interaction and enjoyment needs (Quan-Haase & Young, Reference Quan-Haase and Young2010; Whiting & Williams, Reference Whiting and Williams2013). Thus, we follow Ding et al. to position SMU as a vital online context and explore its moderating effect. Existing studies have discussed the moderating impact of SMU on employee performance, creativity, job satisfaction, work–life conflict, emotional exhaustion, and psychological outcomes. However, these studies have not reached consistent conclusions. These inconsistent findings may be due to the fact that most previous studies have viewed SMU as a single-dimensional construct rather than a multidimensional construct, and the impact of different SMU purposes is not differentiated (Van Zoonen, Verhoeven, & Vliegenthart, Reference Van Zoonen, Verhoeven and Vliegenthart2017). Ali-Hassan, Nevo, and Wade (Reference Ali-Hassan, Nevo and Wade2015) described the three components that comprise SMU, namely SU, CU, and HU. Social use is related to the development of team social networks and the maintenance of team relationships; CU is related to team-sharing activities and knowledge creation; and HU refers to using social media for fun, passing time, relaxing, escaping, and entertainment (Ali-Hassan, Nevo, & Wade, Reference Ali-Hassan, Nevo and Wade2015; Papacharissi & Mendelson, Reference Papacharissi, Mendelson and Papathanassopoulos2011; Quan-Haase & Young, Reference Quan-Haase and Young2010). Therefore, we follow Ding, Liu, and Huang (Reference Ding, Liu and Huang2019) to consider SMU as an IT-related contextual factor and validate its moderator effect, which can serve as a counterweight to performance-centric research (Ali, Wang, & Boekhorst, Reference Ali, Wang and Boekhorst2021).

The moderating role of SU

The SU of social media refers to team members’ contact and interaction with other members, sharing social activities and personal information, expanding their social network, and keeping in touch with existing network members through social media (Ali-Hassan, Nevo, & Wade, Reference Ali-Hassan, Nevo and Wade2015; Papacharissi & Mendelson, Reference Papacharissi, Mendelson and Papathanassopoulos2011). The moderating effect of SU is reflected in that it provides interpersonal and emotional support for team members to implement TCB under SL.

The SU of social media provides a platform for team members to instantly share information, ideas, and resources. Smooth interaction can enhance the effectiveness of team members’ collective awareness of SL tasks, thereby encouraging team members to take proactive action to solve problems (Cao & Ali, Reference Cao and Ali2018). Based on the principle of reciprocal norms of social exchange, a higher level of SU helps team members get more support from the workplace (Goh, Ling, Huang, & Liew, Reference Goh, Ling, Huang and Liew2019), so they will have a sense of obligation to the team and give back to the team. Besides, the SU of social media encourages team members to share personal and general information (Ali, Wang, & Boekhorst, Reference Ali, Wang and Boekhorst2021) and promotes frequent interaction and informal communication among members with similar hobbies (Mcfarland & Ployhart, Reference Mcfarland and Ployhart2015), so that the team can form stronger social ties. Furthermore, under a strong level of SU, SL can better promote communication and collaboration among team members, and this real-time collaboration helps reduce misunderstandings and provide good interpersonal support for team members to implement TCB (Ali-Hassan, Nevo, & Wade, Reference Ali-Hassan, Nevo and Wade2015; Quan-Haase & Young, Reference Quan-Haase and Young2010). Therefore:

The moderating role of CU

The CU of social media provides information and knowledge support for team members to implement TCB under SL (Papacharissi & Mendelson, Reference Papacharissi, Mendelson and Papathanassopoulos2011). Social interaction of personal information promote the sharing of information and knowledge within the team, which will help SL to communicate change goals, visions, and strategies, and promotes understanding and engagement among team members (Ali, Wang, & Boekhorst, Reference Ali, Wang and Boekhorst2021; Ali-Hassan, Nevo, & Wade, Reference Ali-Hassan, Nevo and Wade2015). Thus, team members can obtain information support related to the implementation of TCB. The CU gives a platform for team members to simply and effectively communicate with each other, which will enhance the effect of SL and improve team members’ common perception of SL tasks. Team members who know each other well are more likely to share specialized knowledge and expertise than those who do not (Zhou & Mou, Reference Zhou and Mou2021). Therefore, CU can help team members carry out collaborative knowledge retrieval, sharing, and creation, and help them better understand existing information and knowledge, in order to strengthen the effectiveness of SL and thus build on it to create and change.

Besides, from the view of knowledge complementarity, SL gives team members the ‘leader’ role to encourage them to share information and knowledge (Ali-Hassan, Nevo, & Wade, Reference Ali-Hassan, Nevo and Wade2015). In conjunction with that, CU establishes an instrumental link for knowledge integration, promotes the positioning and mutual understanding of team members’ knowledge, and reduces conflicts and misunderstandings in the process of SL (Sinha, Chiu, & Srinivas, Reference Sinha, Chiu and Srinivas2021). Under a high level of CU, SL will help team members generate more knowledge-sharing behaviors, thus enhancing team members’ cognitive awareness of work situations and cognitive trust in each other, as well as strengthening team members’ transformation and knowledge-creation behaviors (Lyndon, Pandey, & Navare, Reference Lyndon, Pandey and Navare2020). Then the following hypothesis is put forward:

The moderating role of HU

The HU of social media is the social network-based entertainment use of team members, which refers to using social media for fun, passing time, relaxing, escaping, and entertainment (Ali-Hassan, Nevo, & Wade, Reference Ali-Hassan, Nevo and Wade2015). Thus, HU can be a problem when team members experience the fun of performing habitual reward behaviors (Ali-Hassan, Nevo, & Wade, Reference Ali-Hassan, Nevo and Wade2015; Lai & Chong, Reference Lai and Chong2021).

High-level HU will weaken the impact of SL on TCB. That is, for one thing, SL within high-level HU settings leads team members to reduce the working time to complete tasks, thus reducing the work input of assuming the role of ‘leader’ or ‘follower’, which weakens the effectiveness of SL, and thus resulting in the reduction of TCB (Luqman, Cao, Ali, Masood, & Yu, Reference Luqman, Cao, Ali, Masood and Yu2017). In such circumstances, team members may acquire a persistent urge to undervalue long-term risk, overemphasize short-term pleasure, and reduce TCB (Lai & Chong, Reference Lai and Chong2021; Luqman et al., Reference Luqman, Cao, Ali, Masood and Yu2017). For another, SL within high-level HU settings can lead to team member distraction and work interruptions. This effect is particularly significant in knowledge-intensive teams, where even minimal levels of distraction can cause individuals to lose focus on their work, resulting in reduced innovation investment and shifts in their roles as ‘leaders’ or ‘followers’. Consequently, the implementation of TCB is diminished (Ali, Wang, & Boekhorst, Reference Ali, Wang and Boekhorst2021). Then the following hypothesis is put forward:

Participants and procedure

We adopted a team-member matching questionnaire. First, we utilized alumni resources to identify 60 companies operating in a team-based structure across various industries, such as finance, electronics, Internet, services, and manufacturing. Each company formed a self-managing work team consisting of 4–10 members who were responsible for coordinating their task plans. Before formal research, the semi-structured online interviews were conducted with a team representative to gain further insights of their leadership styles and the SMU within the team. In order to ensure the correct matching of the team, we sent a specific questionnaire link belonging to this team to the representative of each team, and then he/she sent this link to other members of his/her team.

After confirming the SL structure and social media usage of the samples, we proceeded to collect a three-wave dataset. In the first wave (T1), demographic variables such as gender, age, educational background, team experience, SL, and social media usage were recorded by team members. The questionnaires were sent to 60 teams in the first wave, resulting in an effective collection of responses from 51 teams. Each valid questionnaire was assigned a number for ease of reference in the second wave. One month later, in the second wave (T2), the previous respondents were contacted and requested to report their input on PIS and EI. Again, we collected valid questionnaires from the same 51 teams. Following a 1-month gap, the third wave (T3) involved sending questionnaires to teams from the first two waves, with 47 teams ultimately submitting their responses on TCB. After removing invalid questionnaires with incomplete or irregular answers, the final effective data consisted of 40 teams and 258 individual members. With the help of alumni, we ensured a 90% response rate for each team because a high response is recommended to conduct network analysis that we use to measure SL (Stanley, Reference Stanley and Katherine1994), thus meeting the needs of calculating network analysis. The average team size in the sample was 6.45. Among the participants, 50.39% were male, 38.76% held a bachelor’s degree, 20.93% held a master’s degree, and 21.32% held a PhD degree.

Measures

The variables measured include SL, PIS, EI, TCB, and SMU. The scales were translated from English into Chinese through Brislin’s (Reference Brislin, Lonner and Berry1986) translation approach. All variables were measured with a 7-point Likert scale (ranging from 1 = strongly disagree to 7 = strongly agree).

SL

To quantify SL among each team, we employ a social network approach established by Mathieu, Kukenberger, D’Innocenzo, and Reilly (Reference Mathieu, Kukenberger, D’Innocenzo and Reilly2015). According to their claim, a greater density is associated with increased SL. To assess this, we requested each team member to rate their teammates on a 7-point Likert scale using a single-item question: ‘To what extent does your team depend on this individual for leadership?’ We then summed up all the ratings and divided them by the total potential number of connections among team members. Equation (1) demonstrates the calculation method for determining the density of SL:

(1)\begin{equation}Density = \frac{S}{N}\left( {N - 1} \right)\end{equation}

where S represents the sum of all leadership ratings given by team members, N denotes the total number of team members, and N(N − 1) indicates the total number of potential ties within the team.

PIS

We used six items developed by Stamper and Masterson (Reference Stamper and Masterson2002) to measure PIS, including ‘I feel strongly that I am part of the team’, ‘My team makes me believe that I belong to this team’, ‘I feel that within my team, I am an “outsider”’. Among these, the third, fourth, and sixth questions are reverse scoring questions, and we used SPSS 22.0 to process them and adopt forward scoring. The Cronbach’s α was 0.842 (see Table 1).

Table 1. Reliability and validity analysis

Note. PIS = perceived insider status; EI = emotional intelligence; TCB = taking charge behavior; SU = social use; CU = cognitive use; HU = hedonic use; SMU = social media use; CR = Composite Reliability; AVE = Average Variance Extracted.