Examining the Implications of Negativity Perceptions for Enterprise Social Media Use

Many organizations fail to optimally benefit from voluntary communication and collaboration tools – e.g. enterprise social media (ESM) – where use depends on workers’ discretionary behaviors. This study explores how ESM use is informed by employees’ perceptions of the content these media convey. Specifically, this paper reports on a survey study ( N = 619) to examine the relationship between negativity perceptions and ESM use, through pro-sharing norms and knowledge-sharing intentions. The findings indicate that negativity perceptions of online communication are met with avoidance responses by organizational members and are associated with low platform usage. This relationship is partially mediated by pro-sharing norms and knowledge-sharing intentions, such that negativity perceptions undermine pro-sharing norms and lower knowl-edge-sharing intention, ultimately reducing ESM use. The findings highlight a potential difference in the underlying psychological mechanisms related to negative media content in organizational environments compared to public and mass media environments. This study integrates media selection literature and media psychology perspectives to study technology adoption and expands our understanding of the potential barriers and drivers of platform use in organizational contexts.

communicative behaviors such as bullying, and the spread of disinformation (Pearce, 2015) and negative content.Despite hopes that social media could serve as a space for open communication and public deliberation, interactions on the platform commonly replicate, reinforce, and reify existing power differentials between individuals and groups (Epstein & Quinn, 2020).Moreover, even when individuals do feel free to communicate their views, they often opt to use social media to vent, criticize, or oppose others (Rim & Song, 2016).The presence, and possibility, of negative, critical, or discordant communication on social media can lead individuals to reduce social media activity or avoid such communicative contexts (Bode et al., 2017;Villi et al., 2021).
Though studies of social media as a platform for expressive emotional communication have predominantly examined communication on publicfacing spaces (Keib et al., 2018), there are compelling reasons to explore whether similar dynamics are present in organizational contexts.This is important because organizations increasingly develop, invest in, and adopt enterprise social media (ESM) to support internal goals associated with communication, networking, and information exchange (Leonardi et al., 2013).Specifically, ESM are defined as: "Web-based platforms that allow workers to (1) communicate messages with specific coworkers or broadcast messages to everyone in the organization; (2) explicitly indicate or implicitly reveal particular coworkers as communication partners; (3) post, edit, and sort text and files linked to themselves or others; and (4) view the messages, connections, text, and files communicated, posted, edited and sorted by anyone else in the organization at any time of their choosing" (Leonardi et al., 2013, p. 2).These enterprise counterparts to public social media have gained traction because of their potential to support employees' and organizations' performance and allow employees to communicate, collaborate, create content, and share knowledge with internal audiences (Yee et al., 2021).
Despite the optimism regarding the possibilities of ESM, research also indicates that as much as 80% of ESM platforms remain underutilized (Rode, 2016).As such, the challenge organizations face is not one of finding, activating, or implementing an ESM platform, but rather motivating use in the form of ongoing and active communication on, and with, the platform.Although some barriers to ESM use have been identified, for example, a lack of trust, lack of time, and required change of behavior (Razmerita et al., 2016), research on what may be deterring employees from adoption, and continued use, of ESM platforms is rather limited.
This study investigates whether and how perceptions of negative communication may inform employees' media choices.Negativity perceptions refer to the extent that users may perceive content on specific media (here ESM) to be predominantly negative.Hence, in this study, we refer to negativity perceptions to indicate a perceived dominance of negative communication in the content available on ESM.Notably, these perceptions can be formed based on the content that is made visible and persists on ESM.Research on media selection in public social media and mass media use has demonstrated how individuals are more attentive to perceptions of negative media content (as opposed to neutral or positive content), ultimately informing media selection (as opposed to media avoidance) (Knobloch-Westerwick et al., 2020;Rozin & Royzman, 2001;Van der Meer et al., 2020;Soroka et al., 2018).We contest this assumption in the context of organizational communication and employees' media choices based upon principles of social exchange theory (Cropanzano & Mitchell, 2005;Wang et al., 2020).In the context of organizations, the perceived costs and benefits of social media use may differ relative to public social media use (Treem et al., 2015), and as a result, negativity perceptions may lead employees to avoid rather than use ESM.Because communication platforms such as ESM may serve as settings that facilitate the sharing of negative communication, and organizations often seek to increase the use of these platforms, it is critical to investigate the relationship between perceptions of negative communication on ESM and ESM use.
By examining motivations and deterrents associated with workers' ESM use, this research contributes to communication and media scholarship in two distinct ways.First, it highlights the importance of considering the ways organizational media are emotionally laden and the relationship this might have to workers' decisions regarding media use.Second, it casts doubt on claims that organizational media such as ESM, and other emerging platforms that facilitate greater visibility of communication among workers, are likely to facilitate more transparent, inclusive, or participatory contexts of communication.Individually and in combination, these contributions extend and expand the study of workers' use, or lack thereof, of communication technologies that are discretionary and potentially highly visible in organizational contexts.

Negative Employee Communication on ESM
There may be several reasons why employees generate negative content on enterprise platforms (Reychav et al., 2019).Research on emotion sharing in organizations suggests that employees often seek out others to share and express their emotions, particularly their negative feelings (Reynolds-Kueny & Shoss, 2020).Employees often use communication technologies to vent negative feelings about work or warn others about organizational issues (Lee & Kim, 2020).Some employees may view ESM as a channel to communicate specific frustrations or share critical personal observations (Holland et al., 2016).Second, negative communication could result from an individual experiencing stress, or simply be related to one's tendency to be pessimistic or discouraging.Reynolds (2007) identified three potential sources of workplace negativity: the self, circumstances, and the organization.Essentially, negativity can be etiologically intrinsic to the individual.For example, individuals may experience subclinical depression or conflictual relationships.Socio-situational negativity may also be stress-induced by circumstances.For instance, negativity can occur when employees find themselves caught between work demands and personal stressors.Finally, organizational-level sources of negativity can stem from poor leadership, which may foster feelings of frustration and negativity.Importantly, individuals may have some, but often limited, agency in eliminating or avoiding negativity in an organizational environment.
Furthermore, beyond these potential reasons for negativity, negative content may also become increasingly visible through ESM.For example, one common goal for the implementation of ESM is to facilitate the discussion and resolution of organizational problems by matching different experts across organizational units (Mäntymäki & Riemer, 2016).However, such a focus on problem-solving may also lead to problem crowding, whereby more problems are surfaced than solved (Haas et al., 2015).This highlighting of negativity may be more likely on a platform like ESM since problems and topics often remain concurrently visible while solutions may be hidden in long threads or individual responses on social media platforms.Hence, users may perceive that they are disproportionally exposed to (negative) communication about difficulties on ESM.

Negativity Perceptions and ESM Use
Media and psychology literature provides two alternative ways of examining the implications of negativity in the media -i.e., negativity bias and negativity avoidance.Importantly, negativity perceptions and negativity bias are conceptually and theoretically distinct in various ways.While negativity perceptions refer to interpretations of content on (enterprise) social media being predominantly negative, negativity bias reflects the idea that "negative events are more salient, potent, dominant in combinations, and generally efficacious than positive events" (Rozin & Royzman, 2001, p. 297).Negativity bias has primarily been researched in the context of individuals' news consumption (Niven, 2001), online reviews (Wu, 2013), and information processing (Chung & Lee, 2020;Knobloch-Westerwick et al., 2020;Lang et al., 2007;Möller et al., 2021).
Typically, the assumption is that negative information triggers more cognitive involvement (Möller et al., 2021), especially at moderate levels of arousal (Chung & Lee, 2020;Lang et al., 2007), and in turn leads to more resource allocation (Yegiyan & Lang, 2010), social sharing behavior (Berger & Milkman, 2012), participation in interactions (Joyce & Kraut, 2006), and reciprocity in communication (Stieglitz & Dang-Xuan, 2013).Overall, it has been demonstrated that negative information garners more attention among media users than positive information, resulting in higher media selection and consumption (Knobloch-Westerwick et al., 2020;Rozin & Royzman, 2001;Van der Meer et al., 2020;Zillmann et al., 2001).Yet, in this study, we argue that negative communication and content on ESM will not have such an appeal to employees but rather leads to lower ESM use.Building on Social exchange theory (SET) as a highly influential paradigm for understanding workplace behaviors (Cropanzano & Mitchell, 2005) and as a family of conceptual models that pertain to the sequential transactions between individuals (Cropanzano et al., 2017), we show how negativity perceptions may reduce ESM use.
At its core, SET posits that individuals engage in behaviors that they perceive as positive and provide more rewards than perceived costs -e.g., the development of trusting, loyal, and mutual commitments (Cropanzano & Mitchell, 2005;Wang et al., 2022).In essence, social exchange theories suggest that individuals exchange resources through a process of reciprocity (Colquitt et al., 2013).When employees perceive communication to have high levels of negativity, for instance, through incivility (Miranda et al., 2020), bullying (Pearce, 2015), or unfavorable content, they may perceive it as providing more costs than rewards and therefore reduce their use of the platform.In the context of knowledge sharing, SET has demonstrated that individuals base their decisions to contribute on the evaluation of anticipated costs and benefits of the exchange (Rode, 2016).Hence, when employees perceive communication on ESM as negative, this may reduce ESM use, as the exchange may be more likely to involve critique or negativity (Ziegele & Reinecke, 2017).
The assumption of ESM avoidance is supported by research on workplace incivility -i.e., low-intensity negative behaviors -suggesting that employees subjected to incivility may try to avoid the instigator and withdraw from work (Miranda et al., 2020;Porath & Pearson, 2012).Similarly, Park and Haun (2018) found that e-mail incivility triggered work withdrawal behaviors.Occupational health research has identified a range of employee responses to negative emotions in the workplace, including withdrawal behaviors, avoidance, substitutive acts, or even leaving the workplace (Shaw et al., 2013).Similarly, research on topic avoidance -i.e., withdrawal from informationdemonstrates that people may avoid topics for a variety of reasons such as protecting self-identity, managing information and relationships, and avoiding conflicts (Afifi & Guerrero, 2000).When negative communication occurs on ESM, other users can often avoid exposure, for instance, by abandoning the platform.In doing so, users may escape uncomfortable situations and restore emotional stability (Li et al., 2020).Hence, individuals might adopt ostrich-strategies and attempt to avoid communicative contexts they view as predominantly negative.Hence, we propose: H1: Negativity perceptions of communication on ESM are negatively associated with ESM use.

Negativity Perceptions and Sharing Norms
Within online communities and organizations, negative content may be at odds with the ideological goals and norms of the group aimed at collaboration and socialization (Glikson & Erez, 2013).Research demonstrates that employees weigh negative information carefully when considering their knowledgesharing behaviors (Connelly et al., 2012).Negativity in the workplace (e.g., (cyber)bullying, public feuds) may reduce perceived benefits and uses of social media through reduced pro-sharing norms.Rather than viewing ESM as collaborative spaces, the platform may become viewed as a source of negativity.Moreover, in an environment that is viewed as more negative rather than pro-sharing, employees may feel less motivated to contribute and share their knowledge and expertise (e.g., Connelly et al., 2012;Wu et al., 2022).As such, the extent to which negativity perceptions may contaminate pro-sharing norms in the organization may contribute to reduced sharing intention (Wu et al., 2022) and reduced use of ESM.
Pro-sharing norms refer to a willingness to value and respond to diversity, openness to conflicting views, and tolerance to failure (Bock et al., 2005).Research has convincingly demonstrated that for group members, the properties of the group, including group norms, need to be inferred from others' and one's own actions (Kankanhalli et al., 2005).For instance, Ghosh et al. (2004) note that in distributed teams, unexpected events and unanticipated problems encountered while working together, shaped and created group norms.
Conversely, knowledge-sharing norms in organizations can partially be derived from the extent to which the organizational climate is considered fair, innovative, and collectively oriented (Bock et al., 2005).In addition, in the context of mediated communication, group members have been demonstrated to infer social norms based on their perceptions of interactions within the group (Postmes et al., 2000).Geber et al. (2019) found that perceptions of communication content determined the social norms such that positive evaluations of risk-taking topics strengthened perceived pro-risk-taking norms, while negative evaluations of risk-taking weakened pro-risk-taking norms.In addition, research suggested that encouraging information motivates users in online communities to share information and set up a pro-sharing norm (Sun et al., 2014).By extension, discouraging information -i.e., focusing on problems and negative communication -would demotivate contributions and decrease pro-sharing norms.Hence, we suggest that when communication is perceived to be predominantly negative, these perceptions will reduce perceived pro-sharing norms.As pro-sharing norms are also important predictors of individual sharing behaviors, we propose that the relationship between negativity perceptions on ESM use is mediated by pro-sharing norms (Bock et al., 2005;Kankanhalli et al., 2005).

Negativity Perceptions and Sharing Intentions
Social media platforms typically encourage emotional self-expression by incentivizing users to regularly update their thoughts, feelings, and experiences within their social networks (Waterloo et al., 2018).However, emotional interactions can have consequences for the intentions of others to contribute, especially in professional contexts (Hashim & Tan, 2015).For instance, people tend to be more inclined to share their knowledge with others toward whom they feel positive emotions (liking, empathy) than people toward whom they have negative emotions (frustration, anger) (Van Den Hooff et al., 2012).In addition, people tend to be more inclined to share when they view their environment more positively compared to those who view their environment more negatively (Yeo & Marquardt, 2015).
Similarly, research found that individuals who perceive that their organizational climate discourages risk-taking are more likely to frame sharing knowledge gathered outside the organization negatively, because those workers are worried about the potential risks of something going wrong (Boh & Wong, 2015).Research on workplace incivility -e.g., making derogatory remarks, profanity, harassment, and other unjust acts -has demonstrated that such negative events reduce knowledge sharing intentions (Hew & Hara, 2007) and may lead to knowledge hiding (Arshad & Ismail, 2018).For instance, Hew and Hara (2007) report that the workers in their sample suggested "I don't want to share what I know when the person who receives my knowledge is rude or argumentative" (p.2321), Drawing on similar arguments we propose that negativity perceptions may signal an environment that is less conducive, or at least one that is less inviting, to share knowledge and information.Hence, we hypothesize that negativity perceptions may reduce knowledge sharing intentions, while sharing intentions reduce ESM use.

H3:
Negativity perceptions of communication on ESM are negatively associated with knowledge-sharing intentions, whereas knowledge-sharing intentions are positively associated with ESM use.

Pro-Sharing Norms and Sharing Intentions
Finally, research on pro-sharing norms has demonstrated the importance of such actions in guiding individuals' behavior and enhancing a climate for knowledge sharing (Hsu & Chang, 2014;Kankanhalli et al., 2005).Hence, building or maintaining pro-sharing norms is important in motivating organizational members to share their knowledge with others (Orlikowski, 1993).It is well established that individual members behave in accordance with group norms, as they anticipate receiving ingroup respect when enacting those behaviors (Pagliaro et al., 2011).In the context of knowledge sharing, behavioral intentions are strongly influenced by group norms related to sharing (Bock et al., 2005).Hence, we hypothesize a serial mediation such that negativity perceptions are negatively related to ESM use, through prosharing norms and knowledge-sharing intentions.

Sample and Procedure
Our study is situated in a large natural resources company headquartered in Europe. 1 The organization uses an ESM platform and expressed an interest in our study to improve their understanding of the role of ESM in their internal processes.Due to our partnership with the case organization, we were able to engage in extensive discussions with key decision-makers prior to the study.The ESM platform was introduced to enhance colleague communication, streamline information flow, and promote collaboration.It aimed to minimize mass e-mails and reduce reliance on a variety of other communication channels such as WhatsApp, SharePoint team sites, and newsletters.On ESM, employees could join specialized channels, create personal profiles, and generate and share content with colleagues across the organization.
All 3,000 office employees received an e-mail invitation to participate in the study, and 619 employees returned our questionnaire, yielding a response rate of 20.5%.The survey included specific questions related to the ESM platform and solicited responses about experiences related to communication, social norms, and individual sharing intentions.Most of the respondents were male (60.9%), which is in line with the overall gender distribution in the company.The average age of the respondents was 42.56 years old (SD = 11.51).Most respondents obtained a university (58.2%) or applied science degree (26.7%).The average tenure was 9.71 years (SD = 11.74), and 16.5% of the respondents were employed in a managerial position.On average, employees reported working 40.96 hours per week (SD = 7.27).

Measures
Table 1 reports all measurement items with corresponding descriptive statistics and factor loadings.Responses to the survey statements were recorded on a seven-point scale ranging from (1) strongly disagree to (7) strongly agree unless indicated otherwise.The reliability and validity statistics are reported in Table 2. Negativity perceptions were measured using three items based on (Van der Meer et al., 2020).A sample statement included: "I think communication on [ESM name] is generally skewed to the negative rather than the positive." Pro-Sharing norms were measured by adopting the four measurement items retained by Kankanhalli et al. (2005).Pro-sharing norms refer to the prevalence of norms that are intended to facilitate knowledge-sharing in the workplace (Orlikowski, 1993).A sample item included: "There is a norm of cooperation in [name of the organization]." Knowledge sharing intentions were probed using five items from Bock et al. (2005) and Hau and colleagues (2013), measuring the intention to share explicit and tacit knowledge.A sample item included: "I will try to share my expertise from my education or training with other organizational members in a more effective way." Enterprise social media use was measured using five items from van Zoonen and Sivunen (2023).The measure includes items reflecting commonly performed activities on enterprise social media platforms, including posting comments and reading other people's posts.Sample items include "I post to [ESM name]."Respondents were asked to indicate the frequency with which they engaged in these behaviors ranging on a seven-point scale anchored from 1 (never) to (7) multiple times per hour.

Analysis
The hypothesized model was examined using Structural Equation Modeling (SEM) in AMOS.First, the validity and reliability of the measures were examined by estimating a first-order confirmatory factor analysis.In the second step of the analysis, we examined the hypotheses by estimating the structural paths.Model fit was assessed by inspecting a series of fit indices.Specifically, we report the Tucker-Lewis Index (TLI) and the Comparative Fit Index (CFI).Values equal to or above .95indicate excellent model fit.In addition, we report the standardized root mean squared residual (SRMR) and the root mean square of approximation (RMSEA).Values equal to, or below ≤ 0.05 and ≤ 0.08, respectively, indicate an excellent model fit.Finally, the χ 2 statistic (cmin/df) is presented.Model parameters and bias-corrected confidence intervals were estimated using bootstrapping (5,000 bootstrap samples).

Measurement Model
A confirmatory factor analysis (CFA) was performed to investigate the validity and reliability of the measurement model.The model fitted the data well: χ 2 /df = 2.92; CFI = 0.95; TLI = 0.94; SRMR = 0.05 and RMSEA = 0.056 (CI: 0.049, 0.063).The results of the CFA indicated that the average variance extracted ranges from .50 to .62 and the maximum shared variance ranged between .05 and .10.Moreover, the square root of the average variance extracted was greater than the inter-construct correlations.These findings indicate that convergent validity and discriminant validity can be assumed.Finally, the results indicate that the composite reliability for all constructs is .83,while the maximum reliability (H) ranges between .84 and  .86.Hence, the measurement model also demonstrates high reliability (see Table 2), justifying further examination of the structural model to test our hypotheses.
Hypothesis 1 reflects the assumption that negativity perceptions are negatively related to ESM use.The results indicate that there is a significant negative association between negativity perceptions and ESM use (B = −.084CI95% [−.147; −.028], p = .005).This result provides support for hypothesis 1.
Hypothesis 2 posits that the relationship between negativity perceptions and ESM use is partly carried by pro-sharing norms, such that negativity perceptions deteriorate the sharing climate in the organization.The results demonstrate that negativity perceptions are detrimental to sharing norms (B = −.272CI95% [−.371; −.179], p = .001),however, sharing norms do not demonstrate a significant direct effect on ESM use (B = .006CI95% [−.046; .057],p = .838).The indirect relationship between negativity perceptions and ESM use through sharing norms is also not significant (B = −.002CI95% [−.016; .012],p = .803).Hence, the results do not provide support for hypothesis 2.
Finally, hypothesis 4 articulates a serial mediation process, where negativity perceptions are expected to reduce sharing norms, which is assumed to reduce ESM use through knowledge-sharing intentions.The direct relationship between negativity perceptions and sharing norms was already established (B = −.272CI95% [−.371; −.179], p = .001),as was the direct relationship between knowledge-sharing intention and ESM use (B = .106CI95% [.027; .186], p = .002).The results further indicate that sharing norms are significantly and positively related to knowledge-sharing intentions (B = .205 CI95% [.123;.301],p = .001).In line with hypothesis 4, the results showed that the indirect relationship between negativity perceptions and ESM use was serially mediated by pro-sharing norms and knowledge-sharing intentions (B = −.006CI95% [−.013; −.002], p = .001).

Discussion
The findings demonstrate that negativity perceptions of employees are negatively associated with ESM usage.Furthermore, the results indicate that perceptions of communication as negative or unfavorable relate to employees' media choices, and this relationship is partly explained by reduced sharing norms and sharing intentions.Using SET as a theoretical framework, the findings indicate that media choices based on negativity perceptions in organizational settings may operate differently compared to mass media and public social media choices.In public settings, negative communication commonly boosts media selection and attention, but in organizational settings negativity may relate to lower media use.Hence, we found that expressive communication for some may deter the pro-sharing norms and knowledge-sharing intentions of others, and, thus, reduce participation on ESM.

Negativity perceptions and ESM use
Extant research has highlighted the potential role of ESM in initiating and expanding communication and collaboration activities among (dispersed) organizational members (Chen et al., 2019).This literature on ESM use has disproportionally highlighted the drivers and potentially positive implications of these technologies (Aboelmaged, 2018), primarily focusing on opportunities for greater knowledge sharing in organizations (Leonardi, 2014).When communication on ESM is assumed to be task or organization-focused, the expectation that greater visibility of workplace communication is beneficial and desirable is logical.Yet little attention has been paid to the affective aspects of communication in this context, despite findings that negative content is commonplace on public social media (Rashidi et al., 2020;Rim & Song, 2016), and a similar pattern has been documented on ESM (Demetis, 2020).
Hence, this study responds to calls for a deeper engagement with the potential "dark sides" of ESM (Baccarella et al., 2018;Moqbel & Kock, 2018;Sun et al., 2021) and its relationship with how these media are used.Employees share problems (and are often explicitly invited to do so), vent frustration, or might engage in negative voice behaviors or workplace incivility on ESM (Baccarella et al., 2018).Our results show that perceiving the communication on ESM to be negative is associated with less ESM use, implying that the assumed benefits of ESM use are not realized.Though our examination focused only on one aspect of emotion, negativity, the finding highlight the need for further research on the presence of emotional expressions in organizational media and the consequences this has for workers' media choices.
More broadly, these findings are counter to expectations based on research on media selection literature (Keene et al., 2019;Keib et al., 2018;Knobloch-Westerwick et al., 2020) and evolutionary psychology (Rozin & Royzman, 2001) that widely herald negativity as a strong predictor of media selection, attention, consumption, and processing.In discussing how individuals process information, Baumeister et al. (2001) observed that "bad is greater than good," highlighting that the impact of bad events is generally greater than the impact of good events.Our findings suggest that in organizational contexts, media users may refrain from using ESM when predominantly confronted with negative content or the potential threats that such communication may convey.There was no evidence that workers were more inclined to view, share, or consume communication on ESM when they perceived communication as predominantly negative.
In line with SET, the findings suggest that employees may engage in behaviors that are more likely to result in rewarding, positive outcomes and reduce costs -i.e., their exposure to negative outcomes.This is in line with research that demonstrated that public social media users often try to avoid negative or strongly worded, or opinionated posts as it may make them feel uncomfortable and uneasy (Rashidi et al., 2020).Moreover, mediated communication of negative emotions is found to be reciprocated by the withdrawal behaviors of other media users (Dai et al., 2020).Hence, in organizational settings, engaging in behaviors employees perceive as positive, and the desire to develop trusting and mutually beneficial commitments (Cropanzano & Mitchell, 2005;Wang et al., 2022), may outweigh the psychological trigger to engage with negative events or communication (Rozin & Royzman, 2001), for instance through ESM use.

Negativity perceptions, sharing norms, and visibility of communication on ESM
Communication research has predominantly treated pro-sharing norms as a moderating or independent factor that is merely present or absent, without considering how perceptions of communication might influence how these norms take shape at the individual level.The SET framework highlights the importance of considering social norms and rules that guide the exchange of information between individuals (Cropanzano & Mitchell, 2005).This study demonstrates that pro-sharing norms are associated with online interactions and, specifically, perceptions of such interactions.We demonstrate that prosharing norms operate not just as an independent or contextual factor (Kankanhalli et al., 2005;Pee & Min, 2017;Wang et al., 2020) but are shaped by perceptions of communication on ESM.These findings support previous literature on organizational members' willingness to contribute to a shared online resource and how this willingness can be shaped by increases in the perceived costs or benefits of such an action (Cress et al., 2006).
This study indicates ways that discretional organizational media (e.g., ESM, collaborative software applications, intranets) that afford greater visibility of communication media may present a double-edged sword: both empowering some individuals to communicate in expressive ways and deterring participation by others who seek to avoid expressive communication for various reasons.Scholars have noted that recent calls for greater transparency among organizations, governments, and public officials often adopt a spurious logic that making more information publicly visible will provide greater accountability and more meaningful evaluation of behaviors (Flyverbom, 2016).This study demonstrates a potential mechanism by which greater visibility of communication in organizations can have the opposite effect of what is intended.
This study establishes that communication being visible on ESM does not necessarily mean that employees perceive that communication positively.Moreover, focusing on visible communication can obscure the choices, views, and behaviors of individuals who decide not to use visible communication platforms (i.e., those who are not visible).When negative communication is particularly visible in an organization, it deters communication from workers who wish to avoid scrutiny, hold minority opinions, or do not feel they have the power to communicate effectively in such a context.This silencing effect, in turn, can have the relative consequence of amplifying or centering negative communication.This contributes to the literature on the circumstances under which employees may be reluctant to voice unpopular views, call attention to inappropriate workplace behaviors, or communicate opinions contrary to majority views or organizational norms (Clair et al., 2019).These silencing effects may be more prominent in contexts where individuals can choose among multiple channels of communication that vary in features and visibility (e.g., ESM use in organizations) (Askay, 2015;Slater, 2007).Although organizational media like ESM potentially expand the visibility of workers' communication greatly, they also can perpetuate and reify power imbalances in organizations that, in effect, render some workers invisible.

Practical Implications
This study was partly motivated by continued reports that many organizations do not obtain the expected returns from their ESM investments (Chen et al., 2019).This is often attributed to low adoption rates of ESM by employees.The findings indicate that negativity perceptions are directly and indirectly contributing to lower adoption rates.Employees can engage in negative voice behaviors, share problems and frustrations, or engage in sharing negative emotions on ESM.As these messages accumulate and create precedence for similar communication, other users may form impressions that negativity and problems are dominating the platform, discouraging further engagement.Similarly, in threads that follow the initial post, solutions to problems may be obscured by responses of employees who recognize similar problems or emotions as the initial poster, thereby contributing to negativity perceptions of observers.Ultimately this could lead employees to abandon the platform.
These findings suggest that organizations may consider moderating these platforms to prevent excessive negative communication on ESM.Many organizations have social media managers within their communication departments for external stakeholder relations or have outsourced such tasks to community managers and social media marketing agencies.Regarding internal stakeholders, moderation would allow organizations to highlight solutions, provide context for problems, and understand underlying structural problems of negative communication.Simple solutions could be to pin good initiatives and solutions in threads, so they do not disappear from timelines.If negativity perceptions prevail, organizations could invite or promote the sharing of achievements and wins.
In addition, organizations may focus on addressing important antecedents of negative communication.For instance, negative online communication by employees is often motivated by the need to vent frustrations or warn others (Gossett & Kiljer, 2006;Lee & Kim, 2020).However, such needs may be more effectively gratified through other, less public communication.For instance, if frustrations can be voiced to supervisors or other organizational members that may directly address the source of employees' frustration, this might result in the necessary changes to improve the situation.In addition, investing in the organization-employee relationship, in general, is important to reduce negative communication on social media.In line with SET, trust, control mutuality, commitment, and satisfaction are found to reduce negative megaphoning (Lee & Kim, 2020).If there is less negative communication on social media, it becomes less likely that negativity will dominate communication perceptions.
The findings also highlight the importance of protecting pro-sharing norms in organizations.Pro-sharing norms are important for individual knowledgesharing intentions and, ultimately, ESM use, but negativity perceptions may chip away at these norms as employees actively avoid such content.In building and maintaining pro-sharing norms, it is important to establish what successful pro-sharing norms look like and to break down these norms into behaviors.Making community-building behaviors more visible and easier to enact helps employees adequately address, and respond to, negative communication on ESM.It offers employees a constructive way to approach such content rather than avoid it, which could, in many cases, be counterproductive.

Limitations and Future Research
Several limitations need to be acknowledged.First, this study relies on crosssectional data collected in one company.This limits our possibilities to draw any causal conclusions about the proposed relationships and limits the generalizability of the findings.Future research is needed to replicate these findings using longitudinal research designs and sampling techniques that include multiple organizations from various industries.Sampling from one organization ensures that our respondents have a similar understanding of the context within which the study is conducted (e.g., organizational culture, platform use, and purpose), allowing meaningful inferences about what drives the use of ESM in that context.However, we acknowledge that our sampling technique may introduce a lack of diversity in the sample (e.g., different platforms and purposes, different organizational structures, and leadership), limiting the generalizability of the findings.Although the impact of leadership, organizational structures and cultures, or different platforms and purposes were not the focus of our study, future research could benefit from adopting a repetitive cross-case analysis to test the robustness of the model presented in this study against such contextual factors.
Second, our findings indicate that negativity perceptions of communication on ESM are more likely associated with avoidance rather than indulgence.However, future studies may explore in more depth which specific mechanisms and conditions may explain whether negative communication may make employees more or less inclined to use ESM.For instance, platform design (e.g., anonymity, thread design) or organizational factors (e.g., culture, goals) may play a role.In addition, it seems worthwhile to explore the extent to which employees are generally tolerant toward, or accepting of, negativity (Aljawarneh & Atan, 2018), the extent to which these perceptions are related to individuals' beliefs of personal accountability (Treem et al., 2015), or person-environment fit, such as congruence between individual values and those communicated on ESM by others (Pee & Min, 2017).These factors may operate as moderators that could explain under which conditions negativity perceptions may trigger more or less ESM use.
Furthermore, future studies could explore the antecedents of negativity perceptions.For instance, in negative workplace relationships Labianca and Brass (2006) highlight the importance of expectation discrepancies.People typically expect interactions at work to be positive and polite, and as such, negative communication stands out.Future research could examine the extent to which discrepancies related to expected and observed communication on ESM may drive communication perceptions.Conversely, future studies could explore the relationship between negativity perceptions and negativity bias in more detail.For instance, it is possible that negativity bias informs perceptions of negativity as people are more prone to negativity in general.Future studies could deploy longitudinal studies with behavioral trace data of ESM to detect the presence or absence of bias and unravel the relationship between negativity perceptions, bias, and platform use.
Overall, the study makes important contributions by identifying negativity perceptions as a barrier to ESM use.We demonstrate that negativity perceptions are more likely met with avoidance strategies (selective exposure) compared to approach strategies (negativity bias) in organizational contexts.Furthermore, rather than treating pro-sharing norms as independent factors that exert an influence on knowledge sharing dynamics, these findings indicate that negativity perceptions may lead to a decrease in prosharing norms, which in turn may reduce individual sharing intentions and platform use.

Note
1.The organizational site mentioned in this study is the same as in Van Zoonen et al. (2022), but the research presented here is based on a separate survey conducted one year later explicitly designed to investigate the implications of negativity perceptions.

Disclosure statement
No potential conflict of interest was reported by the author(s).

Funding
The work was supported by the Academy of Finland [318416].

Table 1 .
Survey items latent constructs measurement model.
Hair et al. (2010)ngs are significant at p < .05bUnitloading indicator constrained to 1. Note: followingHair et al. (2010)factor loadings of .50 or above were considered acceptable.

Table 2 .
Validity and reliability statistics.

Table 3 .
Standardized and unstandardized coefficients for hypothesized regression model.