10th Biennial ACSPRI Social Science Methodology Conference in conjunction with the Australian Social Network Analysis Conference (ASNAC)

Research on online communities shows that horizontal social control contributes to a sense of community that is important for sustaining these spaces. Through such control, users regulate one another’s behavior by identifying or criticizing norm violations. While these practices can strengthen a sense of community, they also raise concerns that heightened sensitivity to others’ judgments may discourage participation. To explore how social control shapes participation in online communities, this study focuses on call-outs as a public form of norm enforcement through which users identify a user’s behavior as a norm violation and invite others to judge it.

FMKorea, one of the largest online communities in South Korea, provides a “Colosseum” category where any user can post call-outs in the politics and current affairs forum. These posts appear alongside ordinary posts, state a user’s username or ID, and cite their posts or comments as evidence of norm violations, allowing others to assess the cited behavior and whether the call-out is justified.

Such call-outs may provide a pathway to influence within online communities. By publicly pointing out perceived norm violations, call-out authors act as informal norm enforcers and then become more visible within the community. Conversely, influential users can easily be call-out authors drawing on their existing positions. To explore this interplay, this study examines longitudinally whether users’ positions in each network influence their subsequent call-out activity (influence process) and whether their level of call-out activity shapes subsequent changes in their ties within each network (selection process).

Furthermore, call-outs may affect users beyond their authors and targets. When users observe those they interact with engaging in call-outs, they may recognize that they, too, could become targets, encouraging caution in posting and commenting. Alternatively, interacting with call-out authors may give these users greater confidence to express their views. This study therefore examines whether ties with call-out authors inhibit or facilitate audience users’ subsequent activity.

We constructed post–comment (net_P) and comment–reply (net_C) networks from non-call-out posts and comments to distinguish users’ influence in attracting attention through posts from their influence in engaging others in discussion. The analysis included 117 users over three waves; tie formations/dissolutions ranged from 209–238/201–245 in net_P and 162–177/117–153 in net_C. “Colosseum” posts were excluded from both networks and used to measure call-out activity as a behavioral variable. A stochastic actor-oriented model (SAOM) analyzed influence and selection processes while accounting for interdependencies between the two networks.

Preliminary findings suggest that higher indegree in the post–comment network significantly increases subsequent call-out activity, which, in turn, significantly increases the likelihood of receiving replies in the comment–reply network. In addition, ties to call-out authors significantly decrease subsequent commenting among audience users. However, the standardized mean net upvote score (upvotes minus downvotes) received per call-out post was not significantly associated with subsequent call-out activity, nor was call-out activity significantly associated with the average call-out activity of users they interacted with in the comment–reply network.

Our findings suggest that call-outs may enable influential users to convert broad attention gained through posts into direct interaction, becoming more deeply embedded in the community’s relational structure. Meanwhile, call-outs may impose participation costs on audience users and constrain discussion. This study deepens our understanding of how horizontal social control interacts with power relations to shape the distribution of influence and patterns of participation in online communities.

Kakaotalk, the most widely used instant messenger in South Korea, allows anyone to create open-chat rooms where up to 4,000 users can converse. Users can search for and anonymously join these rooms by topic. Such rooms may receive hundreds of messages daily, yet only about ten are visible on a single screen. Thus, tracking the conversation and reviewing another participant’s earlier remarks requires additional effort. Despite these communicative constraints, open-chat rooms have not been adequately examined in research on how users form ties through online conversation. Accordingly, this study examines how message characteristics and the frequency of co-participation shape the formation of conversation ties in these rooms.
The formation of conversation ties may be associated with the extent of similarity in the content of two users’ messages. Users judge from message content how far their interests overlap with another's, and the presence of someone with similar interests lowers the burden of addressing a stranger, engaging the tendency to associate with similar others.
Furthermore, the formation of conversation ties may be associated with the similarity of two users’ message lengths. Length may signal the engagement a person brings to the conversation. Long messages elaborate on thoughts and experiences, whereas short ones remain cursory. Users may address participants whose message lengths reflect a level of engagement similar to their own.
Lastly, the formation of conversation ties may be associated with the frequency with which two users participate in the same conversation. Repeated co-participation in the same conversational stream heightens users’ awareness of one another.
To this end, this study uses a multiple regression quadratic assignment procedure (MR-QAP) to examine how message similarity, differences in message length, and the frequency of co-participation are associated with the subsequent formation of conversation ties. A total of 110,515 messages were collected over roughly four months from an open-chat room focused on AI and divided into two periods, T1 and T2. Rapid developments in AI quickly render accumulated knowledge outdated, making this room well suited to observing how users assess one another and form ties from limited cues. A conversation tie was defined as a directed link from each speaker to the immediately preceding speaker. The T2 network, whose nodes were 349 active participants, served as the dependent variable, while predictors were constructed from T1. Message similarity was measured by cosine similarity between users’ messages, length difference by the absolute difference between their mean message lengths, and co-participation by the number of times they joined the same conversational stream. The T1 network and measures of preferential attachment, reciprocity, and transitivity derived from it were included as controls.
Message similarity is expected to promote the formation of conversation ties by revealing shared interests when little other information is available. Message length difference, by contrast, is expected to have a negative effect because a larger difference indicates dissimilar levels of engagement. Co-participation frequency is also expected to have a positive effect because it increases the likelihood of tie formation through greater mutual awareness and more opportunities to respond.
Overall, conversation ties in open-chat rooms are expected to form through immediate responses to message-level cues when participants have little accumulated information about one another. If our expectations are supported, they extend conventional accounts to a setting where the cues that guide attention are not records of past activity but the messages themselves. This study provides a starting point for examining how mechanisms identified in prior research operate differently in open-chat rooms and which other factors may become influential.

In online communities, view counts, upvote counts, and comment counts indicate how users engage with posts. Posts receiving many view counts and upvote counts may be considered supported by other users, whereas posts with many comment counts tend to generate active discussion by prompting additional opinions. Although comment counts can be positively correlated with view or upvote counts, some users may receive high levels of support but limited discussion, while others may generate discussion despite receiving relatively little support. Thus, high support does not necessarily lead to active discussion, and vice versa. Both types of users exhibit high levels of engagement with their posts and can therefore be classified as influencers, but their characteristics differ depending on whether they primarily receive support from other users or encourage additional discussion. Consequently, treating them as a single type of influencer may obscure the distinct influences and roles that each type plays within the community. Accordingly, this study classifies influencers into three types based on whether they meet criteria for the level of support (view counts /upvote counts) and interaction (comment-based conversation networks), and aims to identify how the characteristics of the posts they create differ across these types.
Using data from the Politics/Current Affairs board of FMKorea, one of the largest online communities in South Korea, this study constructs a conversation network based on post–comment and comment–reply relationships. The conversation network constructed from 20,306 posts and 87,162 comments consists of 4,765 nodes and 52,458 edges. Users with high in-eigenvector centrality are classified as “interaction influencers,” while users receiving high levels of support based on view counts and upvote counts are classified as “support influencers.” Users meeting both criteria are classified as “super influencers.” This approach enables a comparison between users who occupy central positions in discussions via posts and comments and those who receive support from the other user. To compare the three influencer groups, we conduct multiple regression quadratic assignment procedure (MRQAP) analyses using two dummy-coded group variables as independent variables. We estimate separate models in which the dependent variable is pairwise similarity in one of four post characteristics: post length, word use, the use of external links, and the number of comments authors leave on their own posts. These analyses examine whether similarities in post characteristics differ by influencer group.
As an expected result of this study, the similarity in word use is expected to decrease significantly among interaction influencers compared to support influencers. The use of common vocabulary, such as popular expressions used within the community, may help influencers receive more support from other users. The similarity in the use of external links is expected to be higher among interaction influencers than among support influencers. Providing information from external sources in posts may encourage users to participate in further discussions through comments. Interaction influencers and super influencers are also expected to have more similar post characteristics to each other than to support influencers.
This study contributes to the literature by distinguishing influencers from two perspectives—levels of support and the formation of interactive discussions—in order to account for their distinct roles their distinct roles, and by identifying how the characteristics of posts created by influencers who receive support differ from those of influencers who lead discussions. From a social perspective, this study provides insight into the different roles that influencers may play in the process of opinion formation in anonymous online communities, such as sharing information from external sources and encouraging interaction.

Previous climate change communication research has examined how news framing shapes attitudes, emotions, policy support, and behavioral intentions. However, less attention has been paid to whether people with similar characteristics evaluate news using the same frame in similar ways and whether the factors associated with this similarity differ across frames. Do conservatives evaluate climate news in ways similar to other conservatives, and liberals in ways similar to other liberals? Or are similar evaluations associated more with how people feel about climate change and whom they trust for political information?

This study examines whether similarity in news quality evaluations is associated with similarities in climate-related emotions (concern, hope, and anger), climate change knowledge (perceived and objective knowledge), and political factors (political ideology and political information-source trust). It also examines whether these associations differ across climate change news frames.

To identify climate change news frames, 7,855 articles published between 2021 and 2024 by six major national newspapers in South Korea were analyzed using semantic network analysis and CONCOR (CONvergence of iterated CORrelations) analysis. Seven frames were identified: Harmful Impacts, Human Touch, Science, Economic Consequence, Responsibility & Accountability, Climate Action, and Power Battle. A web survey of 621 South Korean adults then measured participants’ evaluations of news stories representing each frame. Pairwise similarity matrices were constructed, and separate Multiple Regression Quadratic Assignment Procedure (MRQAP) models were estimated for each frame using the same set of variables.

The findings show that the climate-related emotions associated with similar news quality evaluations differed across frames. Similarity in concern was positively associated with evaluation similarity in Harmful Impacts, Science, Economic Consequence, and Power Battle. Similarity in hope was positively associated with evaluation similarity in Harmful Impacts, Human Touch, Science, and Economic Consequence. Similarity in anger was positively associated with evaluation similarity in Economic Consequence, Responsibility & Accountability, and Climate Action. These results show that different emotions were associated with similar evaluations depending on the frame used to present climate change.

Climate change knowledge showed distinct patterns across frames depending on whether it was subjective or objective. Similarity in perceived knowledge was positively associated with evaluation similarity in Harmful Impacts and Power Battle but negatively associated with it in Climate Action. Similarity in objective knowledge was negatively associated with evaluation similarity only in Economic Consequence. Perceived and objective knowledge were therefore associated with evaluation similarity in different frames and showed no common pattern across all seven frames.

Political factors showed a clearer pattern. Similarity in political ideology was positively associated with evaluation similarity only in Responsibility & Accountability and Power Battle. By contrast, similarity in political information-source trust was positively associated with evaluation similarity across all seven frames. People who had similar levels of trust in political information sources were more likely to evaluate climate news similarly than people who merely shared a similar ideological position.

This study extends framing research by examining which shared audience characteristics are associated with similar evaluations of news using the same frame. The findings suggest that similarity in climate news evaluations is not explained by knowledge or political ideology alone. Climate-related emotions and trust in political information sources were also associated with similar evaluations, with political information-source trust showing an association across all seven frames. Climate journalism should therefore consider not only the factual and scientific information presented in news but also the emotional and trust-based contexts through which audiences interpret it. In fragmented information environments, differences in these contexts may shape whether audiences develop similar standards for evaluating climate news across different frames.