Workplace Negative Gossip Atmosphere and Employees’ Cyberloaﬁng Behaviors: Effects and Mechanisms

. Most previous studies of workplace gossip pay attention to its impacts on the target of gossip, but few focus on how audience of gossip in the workplace may be inﬂuenced when workplace gossip happens. This study adopts audience-centric perspective to understand workplace negative gossip, proposes the construct of “workplace negative gossip atmosphere” (WNGA), and studies how employees’ perception of WNGA inﬂuence their behaviors (cyberloaﬁng). Based on the theory of social information processing and ego depletion theory, this study proposes a dual path model under the “cognitive-emotional” framework. The data analysis software STATA 15.1 is used to analyse 243 employee data collected at 3 time points, and hierarchical regression analyses are conducted to to examine the model. Results show that: 1) WNGA positively inﬂuences employees’ cyberloaﬁng behaviors; 2) Employees’ cognitive trust and emotional exhaustion play mediating roles between WNGA and cyberloaﬁng behaviors, respectively. 3) Task interdependence is a moderator that changes both the relationship between WNGA and cognitive trust, and the relationship between WNGA and emotional exhaustion. This study broadens the research thinking in the ﬁeld of workplace negative gossip and cyberloaﬁng, and also o ﬀ ers some management suggestions for organizations.


Introduction
Gossip seems to be an indispensable part of people's social life. On the one hand, the conversation between people inevitably involves others. On the other hand, people need to exchange information and maintain social relations in the process of gossip. Gossip is not only common in people's daily life, but also common in the workplace. Research showed that more than 90% of employees participated in workplace gossip [1] and 15% of the communication during the rest time was identified as workplace negative gossip [2]. Under the special background of the COVID-19 epidemic, workplace people seem to be more sensitive to workplace gossip. Besides, the people who employees spend the most time with are not their family members or partners, but their colleagues. It can be seen that workplace gossip plays an important role in employees' daily conversation and is an important way of interpersonal interaction [3]. Therefore, workplace gossip is a research topic that will never out of date and cannot be ignored in the field of organizational behavior research. Brady et al. (2016) [4] defined workplace negative gossip as informal and negative talks about absent members in the organization. It is believed that workplace negative gossip is more "existential" than workplace positive gossip [5], because the incidence of workplace negative gossip is much higher than that of workplace positive gossip [6]. In recent years, workplace negative gossip has gradually become a hot topic in the field of organizational behavior research. Workplace negative gossip will cause direct physical and mental damage to the gossip target, and lead to a series of negative results. Studies have shown that workplace negative gossip can bring negative emotions to the target of the gossip, resulting in emotional exhaustion and lower subjective well-being [7]. The harm brought by workplace negative gossip to employees will be manifested in their subsequent workplace behaviors. Workplace negative gossip will reduce employees' OCB [8], bootleg innovation behaviors [9], and increase their knowledge hiding behaviors [10] and unethical work behaviors [11].
However, most previous studies in the field of workplace negative gossip focused on the gossip target, but few cared about audience of gossip in the workplace. In fact, in a workplace negative gossip event, apart from the direct victim (the gossip target) being affected, audience of the gossip also have a perception and are affected in the way of direct witness, indirect hearing and even passive participation. For example, Guo et al. (2021) [12] took gossip receivers as the research object and found that workplace negative gossip would reduce their psychological security. Although existing studies have gradually begun to focus on the impacts of workplace negative gossip on different objects, the research remains to be enriched. After reviewing existing literature of workplace negative gossip, focusing on different objects and atmosphere of workplace negative gossip are important appeals made by   [13]. Therefore, this study proposes a new construct of "workplace negative gossip atmosphere" (WNGA), which is used to describe the tolerance inclination towards workplace negative gossip among colleagues and the frequency of workplace negative gossip events. This study attempts to answer an important question, what impacts will audience's perception of WNGA bring to them?
Existing studies have not paid much attention to the influence of workplace negative gossip on workplace deviance. However, previous studies have proved that workplace ostracism, which belongs to workplace "cold violence" together with workplace negative gossip, has an impact on workplace deviance. For example, Li et al. (2019) [14] showed that workplace ostracism affects employees' cyberloafing behaviors through emotional exhaustion. Thus, this study proposes that WNGA affects employees' workplace deviance. Because under high WNGA, employees have the unease that they will also be the target of gossip one day, which will increase their psychological pressure and consume their emotional resources. To cope with the huge psychological pressure and exhaustion of emotional resources, similar to being ostracized, employees are likely to show workplace deviance. In addition, it can be seen that, as a type of workplace deviance in the Internet age [15], cyberloafing is typical and relatively easy to get.
To summarize, an important disadvantage of the current research on workplace negative gossip is that most studies only focus on the feelings of gossip targets. It is increasingly important to explore the impacts of workplace negative gossip on different objects. At the same time, the research topics of workplace negative gossip should keep pace with the times and the research should be closely combined with the new technical environment. In view of the current research situation, this study hopes to enrich the research perspectives of workplace negative gossip and inject vitality into the research in this field.
Therefore, this study focuses on the influence of WNGA on employees' cyberloafing behaviors from the audience-centric perspective. Based on the theory of social informa-tion processing and ego depletion theory, this study proposes a dual path model under the "cognitive-emotional" framework. By analyzing empirical data, this study verifies the impact mechanisms and boundary conditions of WNGA on employees' cyberloafing behaviors.
The rest of this paper is organized as follows. In section 2, relevant literature is reviewed and hypotheses of relationships among WNGA, cognitive trust, emotional exhaustion, cyberloafing behavior and task interdependence are proposed. In section 3, research methodology is proposed, including data collection, empirical models and measures. In section 4, results of data analysis are shown and the hypotheses are verified. In section 5, conclusions, implications, limitations and future directions are discussed.

Workplace Negative Gossip Atmosphere and Cyberloafing Behavior
Cyberloafing behaviors refer to employees' behaviors of using the Internet for non-work purposes during working hours, such as sending and receiving personal emails, watching online videos or chatting with friends [16]. Cyberloafing behaviors consume the staff's resources of completing their work duties, bringing the reduction of production efficiency and performance [17]. However, with the continuous deepening of research, the "recovery view" of cyberloafing [18] has gradually been attached importance by scholars. In this view, cyberloafing behaviors are believed to be able to restore employees' personal resources, thus producing positive effects. No matter as workplace deviance or as choices for employees to recover resources, cyberloafing behaviors need more attention in the Internet age. And in this study, the cyberloafing behavior is explored as an influencing result of WNGA.
People's reputation is built on social networks, and social networks are produced in oral communication between people. Therefore, under high WNGA and with the hidden nature of workplace negative gossip, employees will have concern about personal reputation damage caused by workplace negative gossip.
Based on the theory of social information processing [19], an individual's attitude depends not only on his own needs and goals, but also on the information conveyed by the social environment where he lives. And the interpretation of the information determines his subsequent attitudes and behaviors. As a source of information from the social environment, WNGA brings individuals concern of becoming the gossip target. Employees are likely to view this concern as an uncontrollable source of stress and then choose avoidance strategies. In the Internet age, mobile networks are readily available, which makes cyberloafing an important way for employees to cope with stress.
Based on the theory of ego depletion [20], facing the worry of being gossip targets, employees may devote more personal resources to the concern caused by WNGA, which will lead to exhaustion of their resources. On the one hand, employees' consumption of personal regulation resources decreases their self-control ability, making them easily distracted and unable to resist the temptation of cyberloafing behaviors. On the other hand, in the fear of becoming gossip targets, employees' sense of resource insufficiency deprives them of necessary control over working environment. So, it is hard for employees to act in favor of their colleagues or organization, and they tend to consider their own interests more [21]. In addition, cyberloafing behaviors can bring employees temporary psychological detachment and resource recovery to some extent [18]. Thus, the following hypothesis is proposed: Hypothesis 1. WNGA has a positive impact on employees' cyberloafing behaviors.

The Mediating Role of Cognitive Trust
Mcallister (1995) [22] divided interpersonal trust into cognitive trust and emotional trust. Compared with emotional trust, cognitive trust is a kind of judgment based on rationality.
Mcallister also pointed out that cognitive trust and emotional trust are positively correlated, and cognitive trust comes into existence before emotional trust. Therefore, this study chooses cognitive trust instead of emotional trust. Based on the theory of social information processing, WNGA is an important information source which can affect employees' judgment on the interpersonal interaction environment, affecting their work attitudes and behaviors. High WNGA means that employees often witness workplace negative gossip events among colleagues, and they can easily feel that they are in an unfriendly and insincere environment where colleagues judge and even hurt each other. Cognitive trust is based on rational calculation and thinking, and past experience is often an important source of information for judgment [23]. As a kind of experience, perception of WNGA will reduce cognitive trust among colleagues. Employees cannot get psychological resources and support that should have come from colleagues. Negative expectations and actual feedback cannot reduce employees' work stress, so they are likely to choose the covert and easy-to-get way of cyberloafing to relieve stress. Thus, the following hypothesis is proposed: Hypotheses 2. Cognitive trust plays a mediating role on the relationship between WNGA and cyberloafing behaviors.

The Mediating Role of Emotional Exhaustion
Emotional exhaustion is a state of fatigue caused by excessive use of psychological and emotional resources, which is a response of stress [24]. Studies showed that emotional exhaustion leads to the decrease of employee engagement [25] and the increase of counterproductive work behavior [26]. In the workplace, emotional exhaustion has been proved to be the response of work stress and interpersonal stress [27]. With high perception of WNGA, employees worry that they will become or have become the targets of negative gossip. This kind of worry will bring employees great psychological and interpersonal pressure, and they have to consume their psychological resources to relieve these pressures.
Based on the theory of ego depletion, cyberloafing behaviors are the results of insufficient self-control and self-regulation [15]. High WNGA will damage employees' psychological and emotional resources. And when the limited resources are consumed, individuals will experience emotional exhaustion. As a result, employees don't have enough regulatory resources to resist the temptation of cyberloafing behaviors. In addition, the efforts made by employees during work will also lead to fatigue or physiological arousal. If employees are no longer exposed to work-related things, the load response will be reversed and recovery will occur. Cyberloafing behaviors can help employees recover psychological resources by allowing employees to take a break from work temporarily. Thus, the following hypothesis is proposed: Hypotheses 3. Emotional exhaustion plays a mediating role on the relationship between WNGA and cyberloafing behaviors.

The Moderating Effect of Task Interdependence
Task interdependence refers to the degree of resource sharing, information exchange and behavior dependence among team members when finishing work tasks [28]. Task interdependence makes the accomplishment of individual task depend on cooperation with other 4 E3S Web of Conferences 409, 04010 (2023) ICMSEM 2023 members. Few studies have focused on the effects of task interdependence on the relationship between atmosphere and individual variables. To measure the different degrees of task interdependence perceived by individuals more accurately, this study regards task interdependence as an individual level variable [29]. High task interdependence means strong interaction between employees' work and requires effective coordination among employees. Therefore, task interdependence demands the harmony of employees' interpersonal relationships in the workplace. In an environment of low task interdependence, the completion of employees' work tasks is not highly related to their colleagues. So, the requirement for maintaining harmonious relationships is not high, and WNGA has limited influence on employees' cognitive trust and emotional exhaustion. However, in an environment of high task interdependence, the completion of employees' work tasks is highly related to other colleagues, requiring high interpersonal harmony. Employees may be greatly concerned about the damage of personal reputation and interpersonal relationships caused by WNGA. As a result, the response to the decline of cognitive trust and the increase of emotional exhaustion caused by WNGA will be stronger. Thus, the following hypotheses are proposed: Hypotheses 4a. Task interdependence plays a moderating role in the relationship between WNGA and cognitive trust; Hypotheses 4b. Task interdependence plays a moderating role in the relationship between WNGA and emotional exhaustion.
To sum up, the theoretical model of this study is shown in figure 1: Hypothesis 4a: Task interdependence plays a moderating role in the relationship between WNGA and cognitive trust; Hypothesis 4b: Task interdependence plays a moderating role in the relationship between WNGA and emotional exhaustion.
To sum up, the theoretical model of this study is shown in

Samples and Procedures
The samples of this study came from IT, financial, service and other industries. In order to avoid the common method deviation, questionnaire data were obtained at three time points. Individuals in the team were sent questionnaires and personal data were matched according to their numbers. At the first time point, questionnaires of demographic information (age, gender, education level, employee type, team size, tenure and marital status), WNGA and task interdependence were distributed to employees. At the second point, questionnaires of cognitive trust and emotional exhaustion were given. At the third time point, the questionnaire of cyberloafing behavior was distributed, and the interval of each time point was 2 months. After excluding invalid samples, 243 valid questionnaires were obtained. As shown in Table 1

Samples and Procedures
The samples of this study came from IT, financial, service and other industries. In order to avoid the common method deviation, questionnaire data were obtained at three time points. Individuals in the team were sent questionnaires and personal data were matched according to their numbers. At the first time point, questionnaires of demographic information (age, gender, education level, employee type, team size, tenure and marital status), WNGA and task interdependence were distributed to employees. At the second point, questionnaires of cognitive trust and emotional exhaustion were given. At the third time point, the questionnaire of cyberloafing behavior was distributed, and the interval of each time point was 2 months. After excluding invalid samples, 243 valid questionnaires were obtained. As shown in

Empirical Models
The empirical models of the study are as follows. Model 1 and Model 2 are models of the two mediating mechanisms; Model 3 and Model 4 are models of the two moderating effects. Model 1: the mediating effect of cognitive trust between WNGA and cyberloafing behavior Model 2: the mediating effect of emotional exhaustion between WNGA and cyberloafing behavior Model 3: the moderating effect of task interdependence on the relationship between WNGA and cognitive trust Model 4: the moderating effect of task interdependence on the relationship between WNGA and emotional exhaustion EE = β 04 + β 14 × WNGA + β 32 × T I + β 42 × (WNGA × T I) + ε 4 (4) Note: CT represents cognitive trust; EE represents emotional exhaustion; CB represents cyberloafing behavior; TI represents task interdependence.

Measures
To ensure the reliability and validity of the measurement instruments, the measurement scales selected in this study were all from mature scales in high-quality journals. Items were translated and revised appropriately according to specific situations. Items of WNGA, cognitive trust and emotional exhaustion were scored on five-point Likert scales (1=strongly disagree, 5=strongly agree). Items of cyberloafing behavior and task interdependence were scored on six-point Likert scales (1=strongly disagree, 6=strongly agree).  [31]. The scale consists of 3 items, such as "My work makes me feel depressed". (4) Cyberloafing Behavior Cyberloafing behavior was measured by referring the scale developed by Lim & Vivien (2002) [16]. The scale consists of 11 items, such as "I read sports-related websites". (5) Task Interdependence Task interdependence was measured by referring the scale developed by Kiggundu (1983) [32]. The scale consists of 6 items, such as "What I do at work affects the work of my colleagues". (6) Control Variables Besides the commonly used demographic variables (age, gender and education level), this study also controlled the variables that may affect WNGA and cyberloafing behavior, such as employee type, team size, tenure and marital status.
Cronbach's alpha is the most commonly used internal reliability index in sociology and psychology. It is an internal consistency coefficient to evaluate the consistency among the scores of each item in the scale according to a certain formula. Internal reliability is used to determine whether a group of items in the scale measure the same concept. Generally, the value of Cronbach's alpha is between 0 and 1. A high value of Cronbach's alpha means high reliability of the scale. In the study of organizational behavior, the reliability of the scale is considered to be good when Cronbach's alpha reaches 0.70.
Values of Cronbach's alpha of the scales used in this study are shown in table 2. Results show that Cronbach's alpha of each scale is greater than 0.70, indicating good internal consistency of each scale.   Table 4 shows the means, standard deviations and correlations of variables in this study. WNGA was positively correlated with cyberloafing behavior (r = 0.20, p < 0.001) and emotional exhaustion (r = 0.54, p < 0.001), and negatively correlated with cognitive trust (r = -0.34, p < 0.001). Emotional exhaustion was positively correlated with cyberloafing behavior (r = 0.21, p < 0.001), and cognitive trust was negatively correlated with cyberloafing behavior (r = -0.15, p < 0.01). The results of descriptive statistical analysis provided preliminary support for the hypotheses of this paper.

Hypothesis Testing
A series of hierarchical regression analyses were conducted to examine the hypotheses. The regression results are shown in table 5 and table 7.
According to M3, after adding the variable of cognitive trust to M2, cognitive trust was significantly negatively associated with cyberloafing behavior (β = -0.220, p < 0.05). And the direct impact of WNGA on cyberloafing behavior was weakened (β = 0.138, p < 0.05), indicating that WNGA affected cyberloafing behavior through the mediation of cognitive trust. Thus, hypothesis 2 was supported. Similarly, according to M4, after adding the variable of emotional exhaustion to M2, emotional exhaustion was significantly positively associated with cyberloafing behavior (β = 0.152, p < 0.05). And the direct impact of WNGA on cyberloafing behavior was weakened, or even insignificant (β = 0.093, p > 0.05), indicating that WNGA affected cyberloafing behavior through the mediation of emotional exhaustion. Thus, hypothesis 3 was supported.
Further, the Bootstrap mediation tests were also conducted to support hypothesis 2 and hypothesis 3. As shown in table 6, the confidence intervals of the indirect effects of two mediating paths are [0.000, 0.090] and [0.013, 0.166] respectively. Both of them didn't contain 0, which also verified the mediating effects of cognitive trust and emotional exhaustion. Again, hypotheses 2 and 3 were supported. (

3) Moderating Effect
As shown in table 7, according to M8 and M10, the interaction (WNGA × TI) was significantly positively associated with cognitive trust (β = 0.133, p p < 0.001) and emotional exhaustion (β = 0.205, p < 0.01). It indicated that task interdependence played positive moderating roles both on the relationship between WNGA and cognitive trust, and the relationship between WNGA and emotional exhaustion. Thus, hypothesis 4a and 4b were supported.  In order to describe the moderating effects of task interdependence more intuitively, two moderating effect figures ( figure 2 and figure 3) were drawn. It can be intuitively obtained that with the increase of WNGA, the decrease of cognitive trust under high task interdependence was greater than that under low task interdependence; the increase of emotional exhaustion under high task interdependence was greater than that under low task interdependence. These results also supported hypothesis 4a and 4b.

Discussion
Based on the theory of social information processing and ego depletion the individuals in a team to explore the impact mechanisms of WNGA on employee The results show that WNGA has a significant positive impact on employees' c cognitive trust and emotional exhaustion play mediating roles in cognitive and em In addition, task interdependence is a moderator that change both the relation cognitive trust, and the relationship between WNGA and emotional exhaustion.

Theoretical Implications
First of all, this study changes the research object perspective in the field of Most of the existing researches on workplace gossip focus on the gossip target their attitudes and behaviors. However, this study shifts from recipient-centri centric perspective and studies how WNGA affects audience' behaviors. Specifi the influence of WNGA on employees' cyberloafing behaviors, and explores its boundary conditions.
In addition, this study makes a further attempt and exploration in the field of research. This study proposes a brand-new construct of WNGA to describe the to workplace negative gossip among colleagues and the frequency of workplace n construct focuses on the individual perception of team atmosphere, captures the feelings, and effectively explains the "cognitive-emotional" reaction and behavi Finally, this study enriches the understanding of the antecedents of cyberl increasingly accessibility of Internet, the study of cyberloafing behavior's an important theoretical and practical significance. Based on the theory of social i ego depletion theory, this study explores the impact mechanisms of WNGA under the "cognitive-emotional" framework, which is of great contribution to th behaviors.  f social information processing and ego depletion theory, this study investigated plore the impact mechanisms of WNGA on employees' cyberloafing behaviors. GA has a significant positive impact on employees' cyberloafing behaviors, and al exhaustion play mediating roles in cognitive and emotional paths respectively. ndence is a moderator that change both the relationship between WNGA and ationship between WNGA and emotional exhaustion.
ications changes the research object perspective in the field of workplace gossip research. rches on workplace gossip focus on the gossip targets and study the changes in ors. However, this study shifts from recipient-centric perspective to audiencedies how WNGA affects audience' behaviors. Specifically, this study focuses on n employees' cyberloafing behaviors, and explores its different mechanisms and makes a further attempt and exploration in the field of workplace negative gossip ses a brand-new construct of WNGA to describe the tolerance inclination towards among colleagues and the frequency of workplace negative gossip events. This behaviors. Specifically, this study focuses on the influence of WNGA on employees' cyberloafing behaviors, and explores its different mechanisms and boundary conditions.
In addition, this study makes a further attempt and exploration in the field of workplace negative gossip research. This study proposes a brand-new construct of WNGA to describe the tolerance inclination towards workplace negative gossip among colleagues and the frequency of workplace negative gossip events. This construct focuses on the individual perception of team atmosphere, captures the individual's differentiated feelings, and effectively explains the "cognitive-emotional" reaction and behaviors of employees.
Finally, this study enriches the understanding of the antecedents of cyberloafing behaviors. With the increasingly accessibility of Internet, the study of cyberloafing behavior's antecedents has increasingly important theoretical and practical significance. Based on the theory of social information processing and ego depletion theory, this study explores the impact mechanisms of WNGA on cyberloafing behaviors under the "cognitive-emotional" framework, which is of great contribution to the research of cyberloafing behaviors.

Practical Implications
Firstly, to reduce employees' cyberloafing behaviors, a harmonious and effective information sharing platform or channel should be established in the organization. When there is a harmonious and effective information sharing platform or channel in the workplace, the topics of gossip will naturally decrease, and WNGA will also decrease. Secondly, the organization does not have to be allergic to employees' cyberloafing behaviors. Cyberloafing behaviors can help employees cope with negative environment or stress in the workplace to some extent, thus improving the effectiveness of their working hours. Thirdly, the organization should provide employees with more resources and necessary counseling. It matters a lot to ensure that employees have other channels for the recovery of psychological resources that are more conducive to the organization. Finally, in a team with high task interdependence, more attention should be paid to the creation and control of team gossip atmosphere.

Limitations and Future Directions
There are still some limitations in this study which need to be further explored in the future. First, there was no multi-source research design for the collection of sample data in this study, and all variable data were reported by individuals. Therefore, in the future, multi-source questionnaire design should be considered to obtain data. In addition, this study only considered the influence of workplace negative gossip atmosphere, ignoring the roles of positive gossip constructs such as workplace positive gossip and workplace positive gossip atmosphere. Future research could consider the effects of workplace positive gossip constructs. Finally, this study only chose cognitive trust and emotional exhaustion as mediating variables. Future studies may consider other mediating variables or influence paths to better uncover the influence mechanism of WNGA on employees' cyberloafing behaviors.
This study gratefully acknowledges financial support from the National Natural Science Foundation of China (Nos.71872117).