Intergroup Communication in Online Forums: The Effect of Group Identification on Online Support Provision

Abstract

Guided by the frameworks of the social identity model of deindividuation effects (SIDE) and the intergroup sensitivity effect, this experiment examined whether and how group identification manifested in user avatars, along with the supportiveness of others’ comments to a support-seeking post, may affect people’s support provision online. Participants were given the option to reply to a support-seeking post in an interactive forum. Participants’ responses were coded for verbal person-centeredness (VPC) and action-focused supportiveness. The SIDE model received some support. The results suggested that more participants replied to a support-seeking message posted by an in-group than by an out-group. Participants provided higher levels of action-focused supportiveness after viewing supportive comments from in-groups than from out-groups. They also provided messages with lower levels of VPC when exposed to unsupportive comments from in-groups than from out-groups. The intergroup sensitivity effect did not receive strong support. The implications of the study are discussed.

Keywords

intergroup communication online forums support provision the SIDE model the intergroup sensitivity effect

Social support exchange, as an essential form of human communication, plays an important role in people’s physical and psychological well-being (MacGeorge, Feng, & Burleson, 2011; Rains & Young, 2009; Wright, 2002). The Internet makes it possible for people to engage in supportive communication online, where they can disclose problems, share experiences, and exchange comfort and suggestions (High, Oeldorf-Hirsch, & Bellur, 2014; Rains & Keating, 2015; Walther & Boyd, 2002; Wright & Rains, 2014).

Online supportive communication, in particular, has drawn substantial research interest (W. Wang & Shen, 2017; Wright & Rains, 2014). A large body of literature focuses on support forums that address health concerns and help people with illness (Wright, Rains, & Banas, 2010; Yang, Zhong, Kumar, Chow, & Ouyang, 2017). However, people experience a wide range of stressors beyond health issues. More research is needed to examine online support in nonhealth related contexts (Rains & Wright, 2016). Indeed, many generic forums allow people to exchange support. For example, the college student forum “collegeconfidential” provides students an outlet to discuss problems encountered in their college lives. Some gaming forums also include subsections to let gamers discuss nongame related stressors.

Although people generally provide support to distressed others, unsupportive interactions also exist in online forums (Aakhus & Rumsey, 2010; Gibbs, Kim, & Ki, 2016). Given that people vary in their supportiveness in online forums, it is important to identify factors that may lead to more or less support. The social identity model of deindividuation effects (SIDE; Reicher, Spears, & Postmes, 1995) has the potential to generate insight on this matter as this theory concerns visual anonymity, a key feature in most online forums. The SIDE model claims that visual anonymity may deindividuate users and make their group identification more salient. As a result, people tend to react more positively toward in-groups than out-groups. Research has applied the SIDE model to online discussion forums and suggested that visual representations (e.g., avatar) make salient intergroup differences (Rains, Kenski, Coe, & Harwood, 2017). For instance, intergroup differences based on user profiles have been found on a movie-related virtual community (Ren et al., 2012). Despite a lack of research examining intergroup dynamics in online supportive communication, such cases have been noted in online forums. Taking gender as an example, salient gender identity prompts in-group and out-group categorizations (Palomares, 2008). Online support–seeking posts regarding family and romantic relationships (e.g., house chores, a breakup) could draw different reactions from men and women whose gender identities are made salient in their user profiles. When forum users choose visual representations that speak more to their group identities than their individuating identities, the salient group identification may affect their supportive communication.

Despite its potential to explain the varying quality of support provision in online forums, the SIDE model has yet to be applied to this specific context. Extant literature on online support has focused primarily on individual and interpersonal factors that affect the process of supportive communication (DeAndrea, 2015; Durant, McCray, & Safran, 2012; High et al., 2014; Li, Feng, Li, & Tan, 2015; Spottswood, Walther, Holmstrom, & Ellison, 2013; Wright et al., 2010). It is often overlooked that forums are community-based (Rains & Young, 2009) and may display features of intergroup communication. In this study, we aim to use the SIDE model as a guide to examine how group categorization entailed by visual anonymity in online forums may affect people’s response tendency and support provision to a support seeker. In addition, prior research on the SIDE model has mostly examined people’s perceptual outcomes such as impression formations and behavior intentions. This study strives to extend the application of the SIDE model by explicitly examining people’s supportive communication and capturing the nuances in their actual messages.

Sharing a focus on intergroup communication, the intergroup sensitivity effect suggests that people tolerate in-group criticism more than out-group criticism. This pattern has been observed in intergroup communication but have not been explicitly tested in the context of online communities (Hornsey, Oppes, & Svensson, 2002; Morier, Bryan, & Kasdin, 2013). Online forums take the form of mass-personal communication where others’ comments are visible to subsequent viewers (O’Sullivan, 1999; O’Sullivan & Carr, 2018). We are thus interested in applying the intergroup sensitivity effect to intergroup communication in online support exchange.

In addition to a test of the theoretical frameworks noted above, this study aims to introduce a methodological advancement of examining people’s support provision. Previous research has typically used self-reported measures to assess participants’ intention to provide support to a support seeker (High et al., 2014; Rains & Brunner, 2018). Although this method works fairly well to gauge people’s behavioral intentions, it does not provide a direct assessment of their actual responses (Yun & Park, 2011). To address this limitation, the present study employs an interactive forum that people are able to voluntarily leave and post their responses to a support seeker. By tracking the number of people who post a reply and coding the actual responses, we are able to assess people’s enacted support provision with greater precision.

In the sections that follow, we begin with a literature review on the theoretical frameworks and their applications in online forums for supportive communication. Hypotheses are then proposed and tested with an empirical study.

The SIDE Model and Response to a Support Seeker

The SIDE model (Reicher et al., 1995), built upon the social identity theory (Tajfel & Turner, 1986), explicates the effect of visual anonymity on social identification and intergroup communication. According to the SIDE model, some forms of computer-mediated communication (CMC) afford visual anonymity because communication partners are physically separated and cannot see each other. Individuating cues that represent unique characteristics of a person are mostly filtered out in CMC, prompting people to focus more on salient group identity cues (i.e., existing social groups or ad hoc activity groups) and to differentiate others into in-groups and out-groups. Consequently, people show favoritism toward in-groups and largely conform to the norms of those groups, whereas they perceive out-groups to be less likable and are less affected by out-groups in their attitudes and behaviors.

The SIDE model has received substantial research attention over the past two decades (e.g., Lee, 2004; Walther, DeAndrea, Kim, & Anthony, 2010; Z. Wang, Walther, & Hancock, 2009; Yilmaz & Peña, 2014). For example, Tanis and Postmes (2003) found that in the absence of individuating cues, participants preferred an in-group (i.e., from the same university) to an out-group (i.e., from a different university) collaborator. Lee (2004) manipulated visual anonymity by assigning participants to the same or a different profile picture from ostensible group members. Through a group decision-making task via anonymous CMC, the participants were found to conform more to interactants using the same profile picture than to those using a different picture.

Online forums typically allow users to maintain anonymity. Some forums adopt a universal avatar system so that users can choose their avatars from a collection of default ones. Other forums enable users to create their personalized avatars. Research suggests that some forum users adopt avatars that reveal more group membership cues than individuating cues (Rains et al., 2017). Using content analysis, Rains and colleagues (2017) found that partisans use avatars (e.g., donkey image) to display their differences from out-groups in online forums. The prominent group identification not only prompts viewers to perceive in-group and out-group members differently (Guegan, Moliner, & Buisine, 2015) but also leads to higher attachment to in-groups (Farzan, Dabbish, Kraut, & Postmes, 2011; Ren et al., 2012). Drawing upon the SIDE model, we argue that viewers are likely to form impressions of support seekers based on salient group identities revealed in their avatars and thus to categorize them into in-groups or out-groups. Because in-group support seekers are considered “part of us” (Reicher et al., 1995; Tajfel & Turner, 1986), their issues might be perceived as more personally relevant and thus draw more responses from in-group viewers. In contrast, viewers perceive out-group support seekers to be less relevant and are less motivated to contribute to those posts.

Hypothesis 1 (H1): Viewers are more likely to reply to a support-seeking message posted by an in-group support seeker than by an out-group support seeker.

The SIDE Model and Quality of Online Support

In the context of anonymous online environments, salient group identity cues can foster in-group and out-group identification and thus affect people’s attitudes toward others. Based on the SIDE model (Reicher et al., 1995), people tend to exhibit positive attitudes toward in-groups but perceive out-groups negatively. For example, Yilmaz and Peña (2014) studied intergroup communication in real-time online chatting and categorized participants into two subgroups (i.e., “sphinx” or “pyramid”) based on their birth months. The participants in the same subgroup were assigned the same zodiac signs as their profile pictures. The results showed that people exhibited more positive attitudes toward in-groups and had higher willingness to collaborate with them (i.e., members with the same zodiac sign) compared with out-groups (i.e., members with a different zodiac sign).

Extending these findings to online forums, viewers are expected to have more favorable attitudes toward in-group than out-group support seekers. Favorable attitudes, in turn, may motivate viewers to provide quality support. According to the constructivism, the production of high quality support requires motivation (Burleson, 1990, 2003). Motivation is dependent, in part, on contextual cues and communicator characteristics (MacGeorge et al., 2011). For instance, people tend to have a stronger intention to help a support seeker who is not blameworthy for his or her stressor than someone who is blameworthy (MacGeorge, 2001). Research has also suggested that a person who is perceived more favorably receives higher quality support compared with a less favorable counterpart (Li & Feng, 2015). Therefore, viewers are expected to provide in-group support seekers higher quality support compared with out-group support seekers.

To measure quality of support messages, both emotional and behavioral aspects could be taken into account (Li & Feng, 2015; Rains, Peterson, & Wright, 2015). Verbal person-centeredness (VPC) concerns the extent to which messages address targets’ emotions and feelings (Burleson, 1982). Messages low in this dimension challenge or ignore targets’ distressed feelings, whereas messages high in this dimension explicitly acknowledge and contextualize targets’ feelings. Action-focused supportiveness is the extent to which helpers legitimize or endorse targets’ courses of action (Li & Feng, 2015). Messages low in action-focused supportiveness tend to criticize targets’ behaviors, whereas messages high in this dimension endorse or recommend targets’ courses of action. Prior research has suggested that support messages addressing both emotional and behavioral aspects tend to be more effective than messages solely addressing one aspect (Feng, 2009, 2014). Considering the theoretical and empirical importance of emotional and behavioral support, this study considers people’s enacted support with both aspects. The following hypothesis was proposed.

Hypothesis 2 (H2): Viewers will provide responses with higher levels of (a) VPC and (b) action-focused supportiveness to an in-group support seeker than to an out-group support seeker.

Others’ Comments and Online Support Provision

Online forums are mass-personal to the extent that anyone who comes across a support-seeking post can witness the interpersonal interaction between the support seeker and a commenter (O’Sullivan, 1999; O’Sullivan & Carr, 2018). People form impressions of a support seeker based on not only the person’s self-disclosure but also situational factors such as others’ reactions to the support seeker (Li & Feng, 2015). Accumulated research evidence has demonstrated the impact of online user-generated comments on people’s attitudes and behaviors in various computer-mediated contexts (Edwards & Edwards, 2013; Li, Feng, Chen, & Bell, 2015; Walther et al., 2010). For instance, a study on professor rating sites found that students who read positive reviews of an instructor had a more desirable impression of the instructor and a stronger motivation to learn, as opposed to students who read negative reviews (Edwards & Edwards, 2013). Likewise, a study on physician review sites found that viewers of positive comments about a primary care doctor were more willing to choose the doctor compared with viewers of negative comments (Li, Feng, Chen, & Bell, 2015). In the context of online forums, people have access to limited information about a support seeker (Spottswood et al., 2013). Therefore, contextual cues such as others’ comments to a support seeker may provide meaningful information to shape subsequent viewers’ perceptions of the target and affect their support provision (Li & Feng, 2015). Li and Feng (2015) found that the supportiveness of others’ comments positively affected viewers’ support provision through their liking of a support seeker and their perceptions of public opinion.

In addition to the supportiveness of comments, people may also take into account the characteristics of the commenters to interpret their comments. The most readily available information about a commenter is the person’s profile and posted messages. Similar to support seekers, commenters may reveal more social identity cues than individuating cues in their profiles. Under such circumstances, commenters are likely to be deindividuated and be perceived based more on their group categorization than on their personal characteristics (Reicher et al., 1995). Compared with out-group commenters, the supportiveness of in-group commenters is expected to exert more influence on viewers’ attitudes toward a support seeker. Walther et al.’s (2010) study on YouTube comments provided some support for our speculations. Guided by the SIDE model, Walther et al. (2010) examined how YouTube comments posted under an anti-marijuana video affected viewers’ evaluations of the video. They found that viewers who identified more with YouTube commenters were more affected by the valence of these comments in their attitudes toward the video. Based on findings from prior research, we expect that viewers will identify more with in-group than out-group commenters. As a result, viewers’ responses to a support seeker will conform more to the supportiveness of in-group comments than to the supportiveness of out-group comments. The following hypothesis was proposed.

Hypothesis 3 (H3): Viewers will conform more to in-group commenters than to out-group commenters in their support toward a support seeker.

Hypothesis 3a (H3a): Specifically, viewers who read supportive comments from in-groups will provide responses with higher levels of VPC and action-focused supportiveness than those who read supportive comments from out-groups.

Hypothesis 3b (H3b): Specifically, viewers who read unsupportive comments from in-groups will provide responses with lower levels of VPC and action-focused supportiveness than those who read unsupportive comments from out-groups.

Intergroup Sensitivity Effect

Like the SIDE model, the intergroup sensitivity effect (Hornsey & Imani, 2004; Hornsey et al., 2002) concerns intergroup communication and sheds light on the impact of supportive/unsupportive comments from in-groups/out-groups on subsequent viewers’ support provision. The intergroup sensitivity effect posits that when an in-group member is criticized, people would respond more negatively to the criticism made by an out-group than by an in-group (Hornsey & Imani, 2004; Hornsey et al., 2002). Depending in part on commenters’ in-group/out-group identification, viewers may have different interpretations of unsupportive comments faced by an in-group support seeker, which in turn affect their own support provision.

Research has demonstrated the intergroup sensitivity effect in various contexts of intergroup communication (Hornsey & Imani, 2004; Hornsey et al., 2002; Morier et al., 2013). Through a series of studies, Hornsey and colleagues (Hornsey & Imani, 2004; Hornsey et al., 2002) found that Australians reacted less defensively to criticism from Australians (in-group) than to the same comments from foreigners. The same pattern has emerged with college students such that students are more receptive to criticism from others attending the same university than from those attending a different university (Hornsey et al., 2002; Morier et al., 2013).

The anonymous environment of online forums may make users’ group identities more salient, based upon which viewers categorize forum users into in-groups and out-groups (Guegan et al., 2015). According to the intergroup sensitivity effect, when encountering an in-group support seeker, viewers may perceive and react to unsupportive comments differently depending on commenters’ group membership. Viewers are expected to react more defensively to out-group than to in-group criticism of an in-group support seeker. Consequently, they may become more supportive of the in-group support seeker facing out-group criticism compared with in-group criticism. The following hypothesis was thus proposed.

Hypothesis 4 (H4): Viewers will provide responses with higher levels of (a) VPC and (b) action-focused supportiveness to an in-group support seeker who receives unsupportive comments from out-groups than from in-groups.

Method

Participants

Participant recruitment and data collection were conducted at a large Midwestern university in the United States subsequent to obtaining institutional review board (IRB) approval. The participants (N = 513) were recruited through a research participant pool. The participants who failed manipulation check questions were excluded from the data analysis (n = 139, 27.10%), which left 374 valid participants aged 18 to 65 (M = 21.76, SD = 4.82). They participated in the study in exchange for extra credit or a small amount of money. The participants were 62.83% female (n = 235), 35.83% male (n = 134), and 1.34% other gender identity (n = 5). Most participants considered themselves to be Caucasian (75.70%, n = 283), followed by Asian or Asian American (11.5%, n = 43), African American (7.20%, n = 27), and other racial/ethnic groups (5.60%, n = 21). Among the 374 valid cases, 194 participants (51.9%) indicated that they responded to the support seeker. Because this study concerns the supportiveness of textual messages, two responses that only contained emoticons were excluded from the final analysis.

Experimental Design

A lab experiment employed a 2 (support seeker’s group identity: in-group vs. out-group) × 2 (commenters’ group identity: in-group vs. out-group) × 2 (supportiveness of comments: supportive vs. unsupportive) factorial design. This study used an interactive support exchange web page on which the participants were able to view, click, and post. Two post topics relevant to college life were adopted from Li and Feng’s study (2015). In one post, a support seeker complained about a conflicting schedule with a roommate; the poster preferred to study after 11:00 p.m. with the desk light on, whereas the roommate insisted on keeping the light off after 11:00 p.m. The other post involved an internship issue, with the support seeker complaining about doing menial tasks and being treated disrespectfully by a boss.

The avatars of support seeker and commenters on the web page were manipulated in terms of group identity. A poster with an in-group identity used an avatar referring to the university that the participants attend. To increase realism, each poster was assigned a different avatar, such as a university flag or a cake with the university’s logo, to indicate in-group membership. Out-group identities were manipulated by using similar images involving symbols indicating a rival university. Each web page included four different avatars, one for a support seeker and three for commenters.

To manipulate supportiveness of comments, two sets of three comments varying in the degree of VPC and action-focused supportiveness were included. Most comments were selected from real forums and were modified for this study (e.g., derive an unsupportive version based on the supportive version of a comment). Each set included exclusively supportive or unsupportive comments. The two versions of comments were made equivalent in length and reading level. For instance, a supportive comment on the roommate conflict was “Your roommate is acting entitled. They can’t have everything they want. It’s your room too. And 11 pm is way too early to sleep!” The unsupportive equivalent was “You are acting entitled. You can’t have everything you want. It’s their room too. And 11 pm is not too early to sleep at all!”

To ensure that supportive and unsupportive comments differ in their VPC and action-focused supportiveness, four sets of comments varying in supportiveness (i.e., supportive vs. unsupportive) and post topic (i.e., intern vs. roommate) were tested with a pool of 32 college students in a pilot study. They were asked to evaluate VPC and action-focused supportiveness of a set of four comments. VPC was measured with four items created based on several existing measures (High et al., 2014; Li & Feng, 2015; Xu & Burleson, 2001). The items were measured on a 7-point Likert-type scale (1 = strongly disagree and 7 = strongly agree). The items were “The commenters sympathized with the original poster,” “The commenters acknowledged the original poster’s feelings,” “The commenters comforted the original poster,” and “The commenters expressed understanding of the situation that is bothering the original poster.” These items were averaged to create a composite variable (M = 3.72, SD = 2.13, α = .95).

The action-focused supportiveness was measured by four items modified from Li and Feng’s (2015) coding manual as well as Xu and Burleson’s (2001) instruments. The items were “The commenters were critical of the original poster’s behaviors,” “The commenters supported the original poster’s behaviors,” “The commenters recommended actions in line with the original poster’s intention,” and “The commenters endorsed the original poster’s behaviors.” The items were measured on a 7-point Likert-type scale (1 = strongly disagree and 7 =strongly agree) and achieved a satisfactory reliability (M = 3.85, SD = 1.68, α = .83).

Results suggested that supportive comments (VPC: M = 5.82, SD = 0.92; action: M = 5.27, SD = 1.04) were perceived to be higher in both VPC and action-focused supportiveness than unsupportive comments (VPC: M = 1.99, SD = 0.85; action: M = 2.60, SD = 0.98), ps < .001. In addition, VPC (roommate topic: M = 3.73, SD = 2.57; intern topic: M = 3.70, SD = 1.70) and action-focused supportiveness (roommate topic: M = 3.65, SD = 1.88; intern topic: M = 4.03, SD = 1.52) did not differ significantly across topics, ps > .53.

Procedure

Upon arrival, participants were randomly assigned to one experimental condition. They were instructed to thoroughly browse the web page that included a support-seeking post and three comments. It was optional for the participants to create a user name and leave a response after they had finished reading the page. Participants who replied to the support-seeking post were able to see their own responses posted underneath the last comment but were unable to see responses from other participants, a technique embedded in this web page to make sure that all participants were exposed only to the manipulated comments. After reading the page and leaving a response (optional), participants were directed to a Qualtrics survey, which asked about their identification with the support seeker and commenters. Participants were debriefed and thanked after completing the survey.

Measures

Response to a support seeker

This outcome was measured dichotomously by a participant’s actual behavior of posting a reply or not. Those who left a response were coded as responding, whereas those who did not were coded as non-responding. Overall, 194 participants (51.9%) replied to the support seeker and 180 (49.1%) did not.

VPC

A nine-level hierarchy of VPC was adopted to code participants’ responses (Burleson, 1982; Li & Feng, 2015). Each message was coded into one of the nine levels based on the extent to which the message acknowledges and contextualizes a support seeker’s feelings (see Table 1 in the online appendix). Messages low in VPC (Levels 1-3) tend to criticize or ignore a support seeker’s distressed emotions. Messages at the mid-levels (Levels 4-6) implicitly acknowledge a support seeker’s feelings. Messages with high levels of VPC (Levels 7-9) explicitly acknowledge, legitimize, and contextualize a support seeker’s feelings. Each message was coded by its dominant level. After several rounds of training sessions, the first and second authors independently coded 121 overlapping messages (63.02%) and reached a satisfactory reliability (intraclass correlation [ICC] = .83). The second author then coded the remaining data set. Despite spanning across different levels, most messages were located at mid-levels (M = 4.15, SD = 1.66; see Table 1 in the online appendix).

Action-focused supportiveness

Each message was also coded for its behavioral aspect based on a nine-level hierarchy on action-focused supportiveness (Li & Feng, 2015). The supportiveness over a support seeker’s courses of action increases with the level (see Table 2 in the online appendix). Messages at Levels 1 through 4 are opposed to a support seeker’s courses of action. Messages of Level 5 take a neutral stance. Levels 6 through 9 include messages that are in favor of a support seeker’s actions. The same two coders independently coded 121 (63.02%) overlapping cases. After achieving high agreements (ICC = .94), the second coder coded the remaining messages. A frequency table for messages at each level was presented in Online Appendix Table 2 (M = 4.44, SD = 1.28).

Manipulation Checks

After accessing the key dependent variables, each participant was asked one question about their awareness of the support seeker’s group identity. The question was “Based on the information on the forum, the original poster should be a student at . . .” They were provided with six options, including the correct answer, the rival university, two distractor universities, the option of “not paying attention,” and the option of “others.” After answering the manipulation check question on the support seeker’s group identity, participants were asked to evaluate their perceived group identification (Lee, 2004; Walther et al., 2010; Z. Wang et al., 2009) with the group that the support seeker belongs to. A scale of six items (e.g., “I identify myself with this group”; “I see myself belonging to the same group”) on a 7-point Likert-type scale (1 = strongly disagree and 7 = strongly agree) was used. The scores on the six items were averaged to create a composite variable for identification with the support seeker (M = 3.47, SD = 1.71, α = .97).

Participants were also asked two equivalent questions on their awareness of the commenters’ group identity (i.e., “Based on the information on the forum, the commenters should be students at . . .”) and their perceived group identification with the commenters. The scores were averaged to create the composite variable for identification with the commenters (M = 3.65, SD = 1.85, α = .98).

As manipulated, participants identified more with a support seeker from the same school (M = 4.41, SD = 1.58) than one from a rival school (M = 2.54, SD = 1.28), t(356.87) = 12.52, p < .001. Likewise, participants had a stronger group identification with in-group commenters (M = 4.90, SD = 1.53) than with out-group commenters (M = 2.44, SD = 1.23), t(351.27) = 17.04, p < .001.

Perceived realism of the thread was also measured with four items on a 7-point Likert-type scale (e.g., “The entire thread described something that could happen in real life”). Overall, the thread was perceived of a high level of realism (M = 6.10, SD = 0.81, α = .79). The perceived realism did not significantly vary across the factors of a support seeker’s identity, commenters’ identity, or commenters’ supportiveness, ps ⩾ .169. However, participants perceived the topic of roommate (M = 6.26, SD = 0.75) to be more realistic than the topic of internship (M = 5.94, SD = 0.83), F(1, 365) = 14.92, p < .001.

Results

The first hypothesis was tested with all valid cases, including both participants who responded to the support seeker and those who did not. Because H2 through H4 examined people’s enacted support, these hypotheses were only tested with participants who replied. Given that VPC and action-focused supportiveness were not significantly correlated (r = .127, p = .079), they were included as separate dependent variables in subsequent analyses. The factor topic was entered as a covariate. The means and standard deviations across all experimental conditions were presented in Table 1.

Table 1.

Means and Standard Deviations of VPC and Action-Focused Supportiveness.

	In-group support seeker				Out-group support seeker
	In-group commenters		Out-group commenters		In-group commenters		Out-group commenters
	Supportive comments	Unsupportive comments	Supportive comments	Unsupportive comments	Supportive comments	Unsupportive comments	Supportive comments	Unsupportive comments
VPC	4.33 (1.34)	3.73 (1.40)	4.48 (1.89)	4.46 (1.74)	4.41 (1.73)	3.55 (1.50)	3.88 (1.48)	4.32 (1.96)
Action-focused supportiveness	5.25 (1.45)	4.08 (1.13)	4.55 (1.28)	4.21 (1.14)	5.00 (1.22)	4.32 (1.13)	4.42 (1.02)	3.77 (1.34)

Note. VPC = verbal person-centeredness.

H1 examined whether people’s response behavior varied as a function of a support seeker’s in-group/out-group identity. A chi-square test was conducted to examine this hypothesis. In support of H1, the result suggested that significantly more viewers responded to an in-group support seeker compared with an out-group support seeker, χ²(1) = 5.18, p = .023. A contingency table was provided in Table 2.

Table 2.

Viewers’ Response Frequency to an In-Group Versus Out-Group Support Seeker.

Case	In-group support seeker	Out-group support seeker	Total
Response	108	86	194
Non-response	79	101	180
Total	187	187	374

H2 proposed that viewers would be more supportive of an in-group support seeker than of an out-group counterpart. A 2 (support seeker’s group identity) × 2 (commenters’ group identity) × 2 (supportiveness of comments) omnibus test was run twice with VPC and action-focused supportiveness entered as separate dependent variables. Contradictory to the prediction, viewers did not show higher levels of VPC toward an in-group support seeker (M = 4.26, SD = 1.64) than to an out-group support seeker (M = 4.01, SD = 1.68), F(1, 183) = 0.677, p = .41. Moreover, action-focused supportiveness did not significantly differ between an in-group support seeker (M = 4.51, SD = 1.31) and an out-group support seeker (M = 4.34, SD = 1.23), F(1, 183) = 0.775, p = .38. H2 was not supported.

H3 proposed an interaction between supportiveness of comments and commenters’ group identity such that viewers’ support provision would be more affected by in-group comments than by out-group comments. We first examined the interaction effect of commenters’ group identity and supportiveness of comments on viewers’ VPC, F(1, 187) = 3.13, p = .078, and action-focused supportiveness, F(1, 187) = 2.12, p = .147. Tests of simple main effects were then conducted to examine the specific hypotheses. H3a proposed that viewers of in-group supportive comments would provide more support to a support seeker compared with viewers of out-group supportive comments. Contradictory to the prediction, viewers exposed to supportive comments posted by in-groups (M = 4.37, SD = 1.50) and out-groups (M = 4.23, SD = 1.74), t(96) = 0.146, p = .884) did not differ in their VPC. As predicted, viewers of in-group supportive comments (M = 5.15, SD = 1.35) provided messages with higher levels of action-focused supportiveness than viewers of out-group supportive comments (M = 4.49, SD = 1.17), t(96) = 2.829, p = .005, Cohen’s d = .52. Therefore, H3a was partially supported.

H3b predicted that viewers of in-group unsupportive comments would provide less support toward a support seeker compared with viewers of out-group unsupportive comments. Consistent with the prediction, a test of simple main effects suggested that viewers of in-group unsupportive comments (M = 3.65, SD = 1.44) provided messages with significantly lower levels of VPC than viewers of out-group unsupportive comments (M = 4.39, SD = 1.83), t(92) = −2.35, p = .02, Cohen’s d = 0.45. Contradictory to the prediction, viewers of in-group unsupportive comments (M = 4.19, SD = 1.12) did not differ from viewers of out-group unsupportive comments (M = 4.00, SD = 1.25) in their action-focused supportiveness, t(92) = 0.76, p = .45. H3b was partially supported.

H4 was proposed to examine the intergroup sensitivity effect which only concerns the in-group support seekers. To be specific, an in-group support seeker would be less supported if unsupportive comments come from in-groups than from out-groups. To test this hypothesis, we only included respondents assigned to the condition of in-group support seekers. We first examined the interaction effect between supportiveness of comments and commenters’ group identity on respondents’ levels of VPC, and then followed up with tests of simple main effects. The interaction was nonsignificant, F(1, 102) = 0.86, p = .356. Although in the predicted direction, the test of simple main effects suggested that an in-group support seeker who was criticized by in-groups (M = 3.73, SD = 1.40) did not receive messages with significantly lower levels of VPC than those criticized by out-groups (M = 4.46, SD = 1.74), t(48) = −1.93, p = .056. H4a was not supported. H4b concerned viewers’ action-focused supportiveness. The interaction effect was not significant, F(1, 102) = 3.22, p = .076. The test of simple main effects suggested that viewers’ action-focused supportiveness toward an in-group support seeker did not differ as a function of commenters’ group identity when they read unsupportive comments (in-groups: M = 4.08, SD = 1.13; out-groups: M = 4.21, SD = 1.14), t(48) = −0.24, p = .81. Therefore, H4b was not supported.

Discussion

Many forum users limit their disclosure of individuating cues for the sake of privacy protection (Rains & Wright, 2016). Group identities are relatively safe to reveal, as they are less likely to expose people’s individuating identities offline. A scarcity of individuating cues may foster salient group identities in online forums, affect group dynamics, and influence support provision. However, intergroup dynamics have received little research attention in the literature of online support. Drawing upon the SIDE model (Reicher et al., 1995) and the intergroup sensitivity effect (Hornsey & Imani, 2004; Hornsey et al., 2002), this study presents an empirical experiment to examine intergroup communication in online support. The SIDE model received mixed support. More viewers replied to a support-seeking message posted by an in-group than by an out-group. Compared with viewers of out-group supportive comments, viewers of in-group supportive comments provided messages with higher levels of action-focused supportiveness, but not with higher levels of VPC. In addition, as opposed to viewers of out-group unsupportive comments, viewers of in-group unsupportive comments provided messages with lower levels of VPC. But they did not differ in action-focused supportiveness. The intergroup sensitivity effect received little support. Viewers did not provide messages with higher levels of VPC and action-focused supportiveness to an in-group support seeker facing unsupportive comments from out-groups than from in-groups.

This study contributes to the literature of online support by advancing the methodology of examining people’s responses to a support seeker. Prior research often used participants’ self-report to assess their willingness to respond to a support seeker and their intended supportiveness (High et al., 2014). Despite being evidenced a valid measure, people’s behavioral intentions do not equal with their actual behaviors (Yun & Park, 2011). To address this concern, a few recent studies have begun to measure people’s written responses (Feng, Li, & Li, 2016; Li & Feng, 2015; Youngvorst & High, 2018). With a focus on message features, this line of research tends to have all participants leave responses to online support seekers. This approach, however, does not allow researchers to assess people’s voluntary replies. Taking a step further, this study gave participants the autonomy to leave a response in an interactive forum, making possible a direct assessment of participants’ voluntary responding behaviors.

The examination of viewers’ response tendency speaks to the serious challenge of under-contribution faced by online forums (Arguello et al., 2006; Gibbs et al., 2016; Ling et al., 2005). A large portion of online posts remains unanswered (Arguello et al., 2006) and most posts are written by a small number of users (Ling et al., 2005). Surprisingly though, considerably less emphasis has been placed on people’s tendency to post on these forums. Little is known about why some people decide to reply to support-seeking posts whereas others stay silent. This study found that more people replied to an in-group support seeker than to an out-group support seeker. The result, however, needs to be taken with caution. Providing a voluntary response to a support seeker does not necessarily mean that the sender provides quality support. Indeed, viewers were not found to be more supportive of an in-group than an out-group support seeker. The results collectively suggested that people may consider voluntary feedback as an act of support because constructing a message takes time and effort (Oh & LaRose, 2016). While people become motivated to reply to an in-group, they may also feel obligated to be candid toward an in-group. As a result, they may sometimes express honest yet harsh thoughts to an in-group in the hope of helping the person directly confront the problem.

This study used the SIDE model as an overarching theoretical framework to examine the impact of group identification on people’s support provision online. The model gained partial support. Viewers would be more influenced by supportive comments from in-groups than from out-groups on their action-focused supportiveness. Furthermore, compared with unsupportive comments from out-groups, the same comments provided by in-groups led to less person-centered messages. In online forums, users learn about support seekers largely via information voluntarily disclosed in their profiles and posts. When such disclosure is constrained by the limited capacity of online forums to transfer social cues, viewers may use whatever cues available to make sense of the situation before leaving responses (Spottswood et al., 2013). When commenters exhibit salient group identification in their avatars, viewers may draw upon this contextual cue to infer group norms to respond to a support seeker. As a result, they are more likely to conform to in-group norms and deviate from out-group opinions in their support provision. While people generally think support seeking and provision is interpersonal communication, the context of mass-personal communication complicates the communication process by adding factors at an intergroup level. Support provision is not only affected by the act of support seeking but also influenced by factors in a broad scope such as social norms in the community (Gibbs et al., 2016), sizes of the group (Rains & Young, 2009), and group identities of the users. This study reveals that prominent in-group and out-group identification may shape the group norm of an online forum and affect viewers’ support provision.

On the other hand, some findings were inconsistent with the SIDE model. Although the SIDE model suggests that people conform more to in-groups than to out-groups when their group identification is salient, this study did not find viewers’ VPC were more affected by supportive comments from in-groups than from out-groups. Because person-centeredness reflects “an awareness of and adaptation to the affective, subjective, and relational aspects of communication contexts” (Burleson, 1987, p. 305), producing highly person-centered messages may need much personal knowledge of a support seeker and sufficient familiarity with a distressed situation. Given the limited information available about a support seeker online, viewers of in-group supportive comments may not know the support seeker well enough to produce highly person-centered messages. Furthermore, viewers’ action-focused supportiveness did not differ when they were exposed to unsupportive comments produced by in-groups and out-groups. Viewers of both conditions implicitly endorsed behaviors against a support seeker’s intention. When opposing to a support seeker’s intended behaviors, viewers may want to mitigate threat to the target’s autonomy and provide suggestions in an implicit way, leading to a lack of differences on action-focused supportiveness in the two conditions. Future research should examine how different goals and considerations may affect people’s online support provision.

With a similar focus on intergroup communication, the intergroup sensitivity effect looks at people’s reactions to an in-group member who is faced with critics from either in-groups or out-groups. In the context of online forums, this study did not find strong support for the interpersonal sensitivity effect. Viewers did not try to provide significantly more support toward an in-group support seeker who received unsupportive comments from out-groups versus in-groups. The results, albeit not significant, were in the predicted direction. It is likely that the predicted pattern may emerge at a significant level with a stronger manipulation of group identification. In this study, viewers’ group identification with in-group commenters was greater than the mid-point with a small magnitude. Given that the lack of nonverbal cues in anonymous CMC can trigger more prominent group identification, the intergroup sensitivity effect has its potential to explain the intergroup communication in CMC. Future research should continue the effort to test the applicability of the intergroup sensitivity effect in the context of anonymous CMC.

Research has accumulated to demonstrate the benefits brought by online forums for supportive exchange (Rains & Young, 2009; Wright, 2002). The feature of visual anonymity, in particular, provides users a safe environment to exchange social support (Walther & Boyd, 2002). Although the advantages are real and remarkable, they may overshadow potential disadvantages of these forums (Aakhus & Rumsey, 2010), a topic that remains understudied in the online support literature. Forum users vary in the quality of their support provision and may hurt support seekers with unsupportive comments. This study sheds light on the issue that why people may produce less supportive comments in online forums. The finding of lowered VPC in the condition of in-group (vs. out-group) unsupportive comments is consistent with prior research that demonstrated the impact of perceived group norms on uninhibited behaviors (Halpern & Gibbs, 2013). A post hoc test also revealed that viewers of unsupportive comments provided messages with lower levels of action-focused supportiveness compared with viewers of supportive comments. In addition to others’ comments and users’ group identification, future research should examine other contextual factors that may lead to more or less support in online forums.

It is worthwhile to mention that most findings in this study had modest effect sizes, indicating that intergroup factors alone may not be sufficient to explain online support processes and outcomes. Indeed, the process of online support is affected by a confluence of factors at individual, interpersonal, and intergroup levels. Although this study focuses on intergroup dynamics, we acknowledge the importance of individual and interpersonal factors in online support. Support seekers are not passive recipients of others’ responses. Instead, many support seekers actively seek help with a variety of means, such as engaging in self-disclosure and responding to previous comments (Li et al., 2015; Tichon & Shapiro, 2003). How support seekers disclose themselves and interact with others plays a vital role in people’s support provision (Feng et al., 2016; W. Wang & Shen, 2017). In fact, several studies examining the SIDE model in other online contexts have integrated interpersonal factors into intergroup dynamics (Z. Wang et al., 2009; Yilmaz & Peña, 2014). These studies show that group membership and interpersonal communication (e.g., the valence of a communication behavior) concurrently affect people’s perceptions of group members and their subsequent communication. In the context of online forums, interpersonal and intergroup factors are expected to work together to influence online support provision. Identifying concurrent mechanisms will provide valuable insights into research on online support. For example, a support seeker may reveal only group identification in a user avatar but exhibit favorable interpersonal behaviors in communication with commenters (i.e., expressing gratitude to previous commenters). Under such circumstances, both interpersonal and intergroup factors could affect support provision. It will be interesting to learn which factor carries more weight in online support exchange.

Several limitations merit some discussion. First, over a quarter of participants (27.10%) failed manipulation checks on group identification and were excluded from the final analysis. Participants were instructed to read over the entire thread but were not explicitly asked to review the avatars. Therefore, some participants may not pay heed to the avatars. It was also likely that a proportion of participants did not recognize the avatars, especially the out-group ones, even though they carefully reviewed them. In either case, they would fail the manipulation checks. In addition, the fact that many participants failed the manipulation checks resonates with the research finding that people engage in selective processing in face of overwhelming information online (Metzger, Flanagin, & Medders, 2010). An interesting question to examine is whether or not people will be unconsciously affected by avatars on their support provision even when they do not pay attention to these cues. Second, this study manipulated in-group and out-group memberships by using rivalry college identities which are not common in online forums except for those targeting college students. Future research should explore different forms of in-group and out-group identities that may become salient in online forums. Third, this study focused on social identity cues and did not examine individuating cues in user profiles. Given that individuating cues are commonly displayed in user profiles, future research should examine the impact of individuating cues on supportive communication in online forums.

Conclusion

This study offered an empirical test of the SIDE model and the intergroup sensitivity effect in the context of online forums for supportive communication. The results showed some support for the SIDE model as viewers provided messages with higher levels of action-focused supportiveness to a support seeker after reading supportive comments from in-groups compared with out-groups. Viewers also produced messages with lower levels of VPC when exposed to unsupportive comments from in-groups than from out-groups. However, the intergroup sensitivity effect did not receive much support. This study helped us to develop a better understanding on how group identification may affect people’s online support provision. Future research can build upon these findings and incorporate interpersonal factors into the investigation of online support.

Supplemental Material

Online_appendices – Supplemental material for Intergroup Communication in Online Forums: The Effect of Group Identification on Online Support Provision

Supplemental material, Online_appendices for Intergroup Communication in Online Forums: The Effect of Group Identification on Online Support Provision by Siyue Li and Guanjin Zhang in Communication Research

Footnotes

Declaration of Conflicting Interests

The authors declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The authors received no financial support for the research, authorship, and/or publication of this article.

Supplemental Material

Supplemental material for this article is available online.

Author Biographies

Siyue Li (PhD, University of California at Davis) is an assistant professor in the School of Communication at The Ohio State University.

Guanjin Zhang (MA, University of Illinois Urbana–Champaign) is a doctoral student in the School of Communication at The Ohio State University.

References

Aakhus

Rumsey

(2010). Crafting supportive communication online: A communication design analysis of conflict in an online support group. Journal of Applied Communication Research, 38, 65-84. https://doi.org/10.1080/00909880903483581

Arguello

Butler

Joyce

Kraut

Ling

K. S.

Rose

Wang

(2006). Talk to me: foundations for successful individual-group interactions in online communities. In CHI ’06 proceedings of the SIGCHI conference on human factors in computing systems (vol. 2, pp. 959-968). New York, NY: Association for Computing Machinery.

Burleson

B. R.

(1982). The development of comforting communication skills in childhood and adolescence. Child Development, 53, 1578-1588. https://doi.org/10.1111/1467-8624.ep8588469

Burleson

B. R.

(1987). Cognitive complexity. In McCroskey

J. C.

Daly

J. A.

(Eds.), Personality and interpersonal communication (pp. 305-349). Newbury Park, CA: Sage.

Burleson

B. R.

(1990). Comforting as social support: Relational consequences of supportive behaviors. In Duck

(Ed.), Personal relationships and social support (pp. 66-82). London, England: SAGE.

Burleson

B. R.

(2003). Emotional support skills. In Greene

J. O.

Burleson

B. R.

(Eds.), Handbook of communication and social interaction skills (pp. 551-594). Mahwah, NJ: Lawrence Erlbaum.

DeAndrea

D. C.

(2015). Testing the proclaimed affordances of online support groups in a nationally representative sample of adults seeking mental health assistance. Journal of Health Communication, 20, 147-156. https://doi.org/10.1080/10810730.2014.914606

Durant

K. T.

McCray

A. T.

Safran

(2012). Identifying gender-preferred communication styles within online cancer communities: A retrospective, longitudinal analysis. PLoS ONE, 7(11), e0049169. https://doi.org/10.1371/journal.pone.0049169

Edwards

(2013). Computer-mediated word-of-mouth communication: The influence of mixed reviews on student perceptions of instructors and courses. Communication Education, 62, 412-424. https://doi.org/10.1080/03634523.2013.800217

10.

Farzan

Dabbish

L. A.

Kraut

R. E.

Postmes

(2011). Increasing commitment to online communities by designing for social presence. In Proceedings of the ACM 2011 conference on computer supported cooperative work & social computing (pp. 321-330). Association for Computing Machinery.

11.

Feng

(2009). Testing an integrated model of advice giving in supportive interactions. Human Communication Research, 35, 115-129. https://doi.org/10.1111/j.1468-2958.2008.01340.x

12.

Feng

(2014). When should advice be given? Assessing the role of sequential placement of advice in supportive interactions in two cultures. Communication Research, 41, 913-934. https://doi.org/10.1177/0093650212456203

13.

Feng

(2016). Is a profile worth a thousand words? How online support-seeker’s profile features may influence the quality of received support messages. Communication Research, 43, 253-276. https://doi.org/10.1177/0093650213510942

14.

Gibbs

J. L.

Kim

(2016). Investigating the role of control and support mechanisms in members’ sense of virtual community. Communication Research. Advanced online publication. https://doi.org/10.1177/0093650216644023

15.

Guegan

Moliner

Buisine

(2015). Why are online games so self-involving: A social identity analysis of massively multiplayer online role-playing games. European Journal of Social Psychology, 45, 349-355. https://doi.org/10.1002/ejsp.2103

16.

Halpern

Gibbs

(2013). Social media as a catalyst for online deliberation? Exploring the affordances of Facebook and YouTube for political expression. Computers in Human Behavior, 29. 1159-1168. https://doi.org/10.1016/j.chb.2012.10.008

17.

High

A. C.

Oeldorf-Hirsch

Bellur

(2014). Misery rarely gets company: The influence of emotional bandwidth on supportive communication on Facebook. Computers in Human Behavior, 34, 79-88. https://doi.org/10.1016/j.chb.2014.01.037

18.

Hornsey

M. J.

Imani

(2004). Criticizing groups from the inside and the outside: An identity perspective on the intergroup sensitivity effect. Personality and Social Psychology Bulletin, 30, 365-383. https://doi.org/10.1177/0146167203261295

19.

Hornsey

M. J.

Oppes

Svensson

(2002). “It’s OK if we say it, but you can’t”: Responses to intergroup and intragroup criticism. European Journal of Social Psychology, 32, 293-307. https://doi.org/10.1002/ejsp.90

20.

Lee

E.-J.

(2004). Effects of visual representation on social influence in computer-mediated communication. Human Communication Research, 30, 234-259. https://doi.org/10.1111/j.1468-2958.2004.tb00732.x

21.

Feng

(2015). What to say to an online support-seeker? The influence of others’ responses and support-seekers’ replies. Human Communication Research, 41, 303-326. https://doi.org/10.1111/hcre.12055

22.

Feng

Chen

Bell

R. A.

(2015). Physician review websites: Effects of the proportion and position of negative reviews on readers’ willingness to choose the doctor. Journal of Health Communication, 20, 453-461. https://doi.org/10.1080/10810730.2014.977467

23.

Feng

Tan

(2015). How social context cues in online support-seeking influence self-disclosure in support provision. Communication Quarterly, 63, 586-602. https://doi.org/10.1080/01463373.2015.1078389

24.

Ling

Beenen

Ludford

Wang

Chang

. . . Kraut

(2005). Using social psychology to motivate contributions to online communities. Journal of Computer-Mediated Communication, 10. https://doi.org/10.1111/j.1083-6101.2005.tb00273.x

25.

MacGeorge

E. L.

(2001). Support providers’ interaction goals: The influence of attributions and emotions. Communication Monographs, 68, 72-97. https://doi.org/10.1080/03637750128050

26.

MacGeorge

E. L.

Feng

Burleson

B. R.

(2011). Supportive communication. In Knapp

M. L.

Daly

J. A.

(Eds.), Handbook of interpersonal communication (pp. 317-354). Thousand Oaks, CA: SAGE.

27.

Metzger

M. J.

Flanagin

A. J.

Medders

R. B.

(2010). Social and heuristic approaches to credibility evaluation online. Journal of Communication, 60, 413-439. https://doi.org/10.1111/j.1460-2466.2010.01488.x

28.

Morier

Bryan

A. E. B.

Kasdin

(2013). The effects of group identity, group choice, and strength of group identification on intergroup sensitivity. Group Dynamics: Theory, Research, and Practice, 17, 14-29. https://doi.org/10.1037/a0030190

29.

H. J.

LaRose

(2016). Impression management concerns and support-seeking behavior on social network sites. Computers in Human Behavior, 57, 38-47. https://doi.org/10.1016/j.chb.2015.12.005

30.

O’Sullivan

P. B.

(1999). Bridging the mass-interpersonal divide synthesis scholarship in HCR. Human Communication Research, 25, 569-588. https://doi.org/10.1111/j.1468-2958.1999.tb00462.x

31.

O’Sullivan

P. B.

Carr

C. T.

(2018). Masspersonal communication: A model bridging the mass-interpersonal divide. New Media & Society, 20, 1161-1180. https://doi.org/10.1177/1461444816686104

32.

Palomares

N. A.

(2008). Explaining gender-based language use: Effects of gender identity salience on references to emotion and tentative language in intra- and intergroup contexts. Human Communication Research, 34, 263-286. https://doi.org/10.1111/j.1468-2958.2008.00321.x

33.

Rains

S. A.

Brunner

S. R.

(2018). The outcomes of broadcasting self-disclosure using new communication technologies: Responses to disclosure vary across one’s social network. Communication Research, 45, 659-687. https://doi.org/10.1177/0093650215598836

34.

Rains

S. A.

Keating

D. M.

(2015). Health blogging: An examination of the outcomes associated with making public, written disclosures about health. Communication Research, 42, 107-133. https://doi.org/10.1177/0093650212458952

35.

Rains

S. A.

Kenski

Coe

Harwood

(2017). Incivility and political identity on the Internet: Intergroup factors as predictors of incivility in discussions of news online. Journal of Computer-Mediated Communication, 22, 163-178. https://doi.org/10.1111/jcc4.12191

36.

Rains

S. A.

Peterson

E. B.

Wright

K. B.

(2015). Communicating social support in computer-mediated contexts: A meta-analytic review of content analyses examining support messages shared online among individuals coping with illness. Communication Monographs, 82, 403-430. https://doi.org/10.1080/03637751.2015.1019530

37.

Rains

S. A.

Wright

K. B.

(2016). Social support and computer-mediated communication: A state-of-the-art review and agenda for future research. Annals of the International Communication Association, 40, 175-211. https://doi.org/10.1080/23808985.2015.11735260

38.

Rains

S. A.

Young

(2009). A meta-analysis of research on formal computer-mediated support groups: Examining group characteristics and health outcomes. Human Communication Research, 35, 309-336. https://doi.org/10.1111/j.1468-2958.2009.01353.x

39.

Reicher

Spears

Postmes

(1995). A social identity model of deindividuation phenomena. European Review of Social Psychology, 6, 161-198. https://doi.org/10.1080/14792779443000049

40.

Ren

Harper

F. M.

Drenner

Terveen

Kiesler

Riedl

Kraut

R. E.

(2012). Building member attachment in online communities: Applying theories of group identity and interpersonal bonds. MIS Quarterly, 36, 841-864.

41.

Spottswood

E. L.

Walther

J. B.

Holmstrom

A. J.

Ellison

N. B.

(2013). Person-centered emotional support and gender attributions in computer-mediated communication. Human Communication Research, 39, 295-316. https://doi.org/10.1111/hcre.12006

42.

Tajfel

Turner

J. C.

(1986). The social identity theory of intergroup behaviour. In Worchel

Austin

W. G.

(Eds.), Psychology of intergroup relations (2nd ed., pp. 7-24). Chicago, IL: NelsonHall.

43.

Tanis

Postmes

(2003). Social cues and impression formation in CMC. Journal of Communication, 53, 676-693. https://doi.org/10.1111/j.1460-2466.2003.tb02917.x

44.

Tichon

Shapiro

(2003). The process of sharing social support in cyberspace. Cyberpsychology & Behavior, 6, 161-170. https://doi.org/10.1089/109493103321640356

45.

Walther

J. B.

Boyd

(2002). Attraction to computer-mediated social support. In Lin

C. A.

Atkin

(Eds.), Communication technology and society: Audience adoption and uses (pp. 153-188). Cresskill, NJ: Hampton Press.

46.

Walther

J. B.

DeAndrea

Kim

Anthony

J. C.

(2010). The influence of online comments on perceptions of antimarijuana public service announcements on youtube. Human Communication Research, 36, 469-492. https://doi.org/10.1111/j.1468-2958.2010.01384.x

47.

Wang

Shen

(2017). Reciprocal communication in online health support groups: Effects of message format and social support on individual responses to health issues. Media Psychology, 20, 240-266. https://doi.org/10.1080/15213269.2016.1142381

48.

Wang

Walther

J. B.

Hancock

J. T.

(2009). Social identification and interpersonal communication in computer-mediated communication: What you do versus who you are in virtual groups. Human Communication Research, 35, 59-85. https://doi.org/10.1111/j.1468-2958.2008.01338.x

49.

Wright

K. B.

(2002). Social support within an on-line cancer community: An assessment of emotional support, perceptions of advantages and disadvantages, and motives for using the community from a communication perspective. Journal of Applied Communication Research, 30, 195-209. https://doi.org/10.1080/00909880216586

50.

Wright

K. B.

Rains

S. A.

(2014). Weak tie support preference and preferred coping styles as predictors of perceived credibility within health-related computer-mediated support groups. Health Communication, 29, 281-287. https://doi.org/10.1080/10410236.2012.751084

51.

Wright

K. B.

Rains

S. A.

Banas

(2010). Weak-tie support network preference and perceived life stress among participants in health-related, computer-mediated support groups. Journal of Computer-Mediated Communication, 15, 606-624. https://doi.org/10.1111/j.1083-6101.2009.01505.x

52.

Burleson

B. R.

(2001). Effects of sex, culture, and support type on perceptions of spousal social support: An assessment of the “support gap” hypothesis in early marriage. Human Communication Research, 27, 535-566. https://doi.org/10.1111/j.1468-2958.2001.tb00792.x

53.

Yang

Zhong

Kumar

Chow

S.-M.

Ouyang

(2017). Exchanging social support online: A longitudinal social network analysis of irritable bowel syndrome patients’ interactions on a health forum. Journalism & Mass Communication Quarterly, 95, 1033-1057. https://doi.org/10.1177/1077699017729815

54.

Yilmaz

Peña

(2014). The influence of social categories and interpersonal behaviors on future intentions and attitudes to form subgroups in virtual teams. Communication Research, 41, 333-352. https://doi.org/10.1177/0093650212443696

55.

Youngvorst

L. J.

High

A. C.

(2018). “Anyone free to chat?” Using technological features to elicit quality support online. Communication Monographs, 85, 203-223.

56.

Yun

G. W.

Park

S.-Y.

(2011). Selective posting: Willingness to post a message online. Journal of Computer-Mediated Communication, 16, 201-227. https://doi.org/10.1111/j.1083-6101.2010.01533.x

Supplementary Material

Please find the following supplemental material available below.

For Open Access articles published under a Creative Commons License, all supplemental material carries the same license as the article it is associated with.

For non-Open Access articles published, all supplemental material carries a non-exclusive license, and permission requests for re-use of supplemental material or any part of supplemental material shall be sent directly to the copyright owner as specified in the copyright notice associated with the article.

0.00 MB

0.49 MB