Information strategies and affective reactions: How citizens interact with government social media content
First Monday

Information strategies and affective reactions: How citizens interact with government social media content by Nic DePaula and Ersin Dincelli



Abstract
As social media use grows among the general population, becoming a major source of information, government organizations around the world have widely adopted these platforms. While researchers on government social media have acknowledged the potential functions of these technologies for participatory democracy, transparency, and collaboration, we have come to learn that prominent social media sites are also sites for marketing and favorable presentations. Given the diversity of content that may flow through a government social media channel, what kind of content is actually adopted? Moreover, how do users react to these diverse content types? In this empirical study, we collected Facebook posts of local government departments across the United States and categorized each post content using a framework of government social media communication strategies. We then analysed differences in users’ reactions in the form of likes, comments, and shares to distinct types of content. We found a number of statistically significant results, providing some evidence for the effects of communication strategy or content type on measures of user interaction. Our results highlight the affective and symbolic nature of social media but also point to higher levels of interaction to instances where governments themselves engage in dialogue with citizens. We believe these results are important for scholars and proponents of government communication as they complicate the notion that social network sites can be a generative source of rational critical thought and deliberative conversations.

Contents

Introduction
Government communication and information strategies
User interaction with social media content
Methodology
Empirical results
Discussion: The affective bias
Conclusions and future work

 


 

Introduction

As social media use grows among the general population, becoming major sources of information for youth and adult alike (Greenwood, et al., 2016), government organizations around the world are widely adopting the platforms — from local government departments in Albuquerque, New Mexico, to the bureaucratic departments of the European Union Commission. Unlike individual agents that adopt the platforms to browse content and interact with posts as they see fit, however, governments serve more as information disseminators and, of course, are highly regulated bodies, with their approach to the media quickly becoming delineated via policy guidelines and strategies (Klang and Nolin, 2011; Meijer and Thaens, 2013; Mergel, 2016). Earlier work on government adoption of social media conceived of these tools as “transparency” and “anti-corruption tools” (Bertot, et al., 2010) that could be used for participatory “citizen coproduction” and “mass collaboration” (Linders, 2012; Mossberger, et al., 2013). U.S. president Barack Obama, in his first days in office, developed an “open government” memorandum which instructed all federal agencies to become more transparent, participatory, and collaborative with citizens (Obama, 2009). Given the emergent popularity of social media then, they naturally became tools through which transparency, participation, and collaboration goals could be accomplished. Nevertheless, the most common reason cited for actually adopting the tools, at least in one study of the U.S. federal government, was simply fulfilling the top-down mandate, rather than any particular organizational need or citizen demand (Mergel, 2013b). Moreover, as problems with social media now abound (e.g., misinformation, echo chambers, algorithmic control), their potential for collaborative partnerships need to be understood within the logic or tendencies that characterize their actual adoption (e.g., Dijck, 2013; Fuchs, 2017).

Although many a platform may be called “social media”, and our theorization is geared toward understanding “social network sites” and “digital Internet technologies” more generally, in this paper our original empirical analysis is based on Facebook. Facebook is the largest social network site in the world, and it is where most government agencies and departments locate their particular pages or channels. Other channels, such as Instagram and Snapchat, are becoming more popular and prominent in government in the United States (Stone, 2015). Nevertheless, the literature has not given much attention to the multifaceted nature of government communication in any of these platforms, or questioned the paradoxes of their uses. Although previous research has found that “Facebook users are more interested in fun and private issues” rather than serious government information [1], this has not been noted as a problem for e-government theory. There are benefits to technology innovation that are generally supported in the literature (Sandoval-Almazan, et al., 2018) and many purposes can be overwhelmingly considered productive, such as in emergency and crisis management situations (Kavanaugh, et al., 2012; Hagen, et al., 2017). However, the problem is not that e-government promises of an increase in interaction, participation, and collaboration are “more rhetorical” than reality (Sandoval-Almazan and Gil-Garcia, 2012), but that popular social media technologies are systematically driven for emotional responses — not for deliberated or rationalized discussions and interactions.

Our research questions are thus the following: What types of content do governments post on social media? Moreover, how do users react to different types of content? For the empirical component of this study, we first used a typology of government social media communication to categorize the distinct types of content that governments may adopt on Facebook. We retrieved 1,421 posts from various local governmental Facebook pages across the U.S. and coded the distribution of content across selected government department pages. We also developed a set of conservative hypotheses of how users are likely to interact with distinct types of content and carried out statistical tests to help us answer whether or not particular types of content had significant effects on how much users like, share, or comment on a respective piece of information.

We conclude the paper with a discussion on the nature of social media interaction in a governmental context. In light of contemporary literature on the “affective” nature of social media — and the affective nature of politics more broadly — as well as the empirical results of our analysis, we find that counts of Facebook reactions are biased toward affective and symbolic content, thus potentially not suitable for measuring open government and democratic values such as transparency, deliberation, and rational engagement. However, users do respond positively to instances where governments use the platform for online dialogue which seems positive for proponents of citizen-government engagement. However, in this research we did not examine content of government-citizen dialogue, simply the instances where some exchange existed. To understand the extent to which social media can be successful for the political process, including government bureaucratic activity and the quality of citizen participation, requires a deeper investigation into communicated content and meaning of sharing and liking behavior. While we do not address all of the intricacies of this phenomena, this research develops empirical and theoretical understanding of the tendencies that drive social media behavior, inside and outside a governmental context.

 

++++++++++

Government communication and information strategies

Research on government use of social media has in the past focused on the use of social media for democratic goals of transparency, participation, and collaboration (Bertot, et al., 2010; Harrison, et al., 2012; Mergel, 2013a). However, SNS and social media are inherently tools for “identity construction” (Nadkarni and Hofmann, 2012), the exchange of values-based symbolic content (Penney, 2015), acts of “social grooming” (Tufekci, 2008), and marketing (Bellström, et al., 2016). Although certain activities — such as “collaboration” or different types of crowdsourcing functionality — are not generally carried out on social networking sites such as Facebook and Twitter, the communicated content openly posted and shared on these platforms may be coded based on a set of communication or information strategies which we assume to be reflected by the type of content government agencies are posting and sharing. Government social media activity is a relatively regulated activity (Klang and Nolin, 2011; Mergel, 2016). Moreover, individuals working as governmental communicators have specific intentions in the creation of public Facebook posts and other similar acts (Searle, 1969). In this study we therefore analyze and code the “content” or “meaning” of the post via a framework of “types of communication”or “communication strategies” that reflect the context of the communication (i.e., social media) and the intentions that government agencies demonstrate in these contexts. The framework is divided into more specific and more general categories, namely: information provision; input seeking; online dialogue/off-line interaction; and, symbolic presentation, as summarized in DePaula, et al. (2018) and Table 1.

Information provision is understood as the publication of government information and content that fulfils the transparency mandate of government (Bertot, et al., 2010), a defining characteristic of political democracies (Harrison, et al., 2012). In the framework adopted here we distinguish between: (1) descriptive information about governmental programs, events, and policy as operations and events information; and, (2) instructions to keep safety and welfare as public service announcements. Operations and events information is information about the government that could be useful or important for citizens to know, fulfilling at least a minimum level of transparency. Public service announcements, however, are instructions or messages that governments give in order to produce broad responses in society in regards to their safety or awareness about social issues (Shoemaker, 1989). These public directives are related to crisis communication where individuals must be quickly instructed on how to respond to dangerous and difficult situations, one reason for which social media has been adopted (Graham, et al., 2015; Hagen, et al., 2017).

One of the main reasons social media have been categorized as beneficial for governments is that they provide convenient platforms from which agencies and departments can seek citizen information or input (Mergel, 2013b). Input may be obtained by organizations via internal observation of user behavior on a given platform, which cannot be captured from research analysis of public posts published by the departments themselves. However, governments have been observed to post requests for feedback from citizens in the form of “surveys” or “polls” (Waters and Williams, 2011). We refer to this type of communication as citizen information, since the input sought is in reference to how citizens think or behave. Moreover, the literature has noted that government communication also involves more tangible types of input seeking in the form of “fundraising” and “donations” (Golbeck, et al., 2010; Hofmann, et al., 2013), which we refer to as simply fundraising. Input seeking efforts may be characterized as “two-way asymmetric” (DePaula and Dincelli, 2016) since there are instances where it is not clear how the organization itself will respond to the input sought and obtained.

 

Table 1: General and specific categories of government social media content.
Information provisionOperations & events: Content on agency policy, operations, and events.
Public service announcements: Regarding safety, health, and well-being.
Input seekingCitizen information: Requesting feedback, opinion; use of survey or poll.
Fundraising: Asking for donations and contributions to a cause.
Online dialogue/off-line interactionOnline dialogue: Response by agency to user comment on agency post.
Off-line discussion: Off-line event to discuss particular policy issue.
Off-line collaboration: Asking citizens to become active and volunteer.
Symbolic presentationFavorable presentation: Positive imagery, self-referential language of gratitude, and praises of itself.
Political positioning: Taking or expressing a position on a political issue.
Symbolic act: Expressing congratulations, condolences to others. References to holiday, cultural, and historical symbols.
Marketing: Presentation of features with intention to attract individuals to acquire or consume.

 

Social media facilitate interaction via “comments”, by allowing users to add comments to posts and reply to those comments, therefore producing some level of dialogue or conversation. It has been argued that these interactions may improve relationships that governments have with their communities (Bovaird, 2007). However, in practice, the interaction over social media platforms may be limited to simple questions and answers and real-time announcements. Nevertheless, governmental departments, at least to a large extent, allow users to comment on content that governmental organizations post. When governmental organizations themselves reply to these comments, we refer to it as online dialogue. Although the extent of dialogue is relative, for this study we simply measure this variable as binary. Given the liberal-democratic nature of governments discussed in this paper, we also sought to capture instances where governments made references to off-line interaction: both in the form of off-line discussions where citizens met with officials and discussed policy issues; or off-line collaborations, where opportunities were publicized for people to volunteer for a project or assist in some community-based activity.

What we broadly term symbolic presentation are a set of information strategies that reflect both socially symbolic content and self-presentational strategies. The literature on government use of social media has not integrated or discussed at length the nature of symbolic and presentational exchanges on social media. However, outside of government, social media are tools that enable performance via “self-presentation” (boyd and Ellison, 2007; Nadkarni and Hofmann, 2012) and become spaces of “social grooming” where content exchange and the use of language are used to signal social status or develop “information fashions” (Donath, 2007; Tufekci, 2008). Similarly, in this realm social media are used for the adoption and distribution of visual symbols in order to communicate particular social values, such as was the case with the equal-sign and marriage equality campaign on Facebook (Penney, 2015), which we infer reflects the strong visual and image-based characteristic of social media (Treem and Leonardi, 2012).

More specifically, what we refer to as favorable presentation are the strategic attempts to create a portrayal of an individual or organization in order to induce attributions of likability and in general create a positive image of the individuals or organizations involved. This type of communication has also been conceived as a type of “impression management” (DePaula and Dincelli, 2016) and is similar to marketing, which works with an affective dimension in order to present a product and attract people to consume it (Bruner, 1990). What we refer to as political positioning are instances where a political opinion is expressed, which we assume to signal political values and are often associated with a political person. Symbolic acts are communication with reference to other types of social values, such as social status, social affiliations and fashions, as well as “social grooming aids” which tell others you are thinking about them (Donath, 2007). This symbolic content may refer to more universal values or emotions (e.g., feeling good from the good nature of people); or are more ideologically based (e.g., references to one’s own cultural characteristic or history) (Dijck, 2013). Although the exchange of this symbolic content may serve as signals of sociality and politeness, it is unclear whether or not such exchanges strengthen government-citizen partnerships and efforts to collaborate or reflect actual improvements in the government’s relationship with their constituency. Although some characterize these symbolic acts as important part of human evolution (Dunbar, 1996), others characterize these exchanges as “meaningless social chatter” to stimulate peace and contentment [2].

 

++++++++++

User interaction with social media content

The basic ways that users can react or interact with social media content are by: liking, sharing, and/or commenting. Studies in e-government and government information management have explored how users respond to content in the form of likes, shares, and comments (Hofmann, et al., 2013; Bonsón, et al., 2015; Lappas, et al., 2017). These forms of social media reactions are often understood as measures of “success” (Mergel, 2013a) or “effectiveness” (Lappas, et al., 2017) in a governmental context. However, researchers on social media more broadly have noted that social media reactions and interactions in general are about the “affective” or emotional response that content brings (Grusin, 2010; Papacharissi and de Fatima Oliveira, 2012; Paasonen, 2016; Tettegah, 2016). As a case in point, in January 2016, Facebook made available as part of its platform new possible reactions to posts, all of which were emotionally focused (“sad”, “angry”, “wow”, and “happy”). Given that part of the data for this study did not include all of these new reactions, we limit this study to the traditional three reactions.

Studies on government social media often distinguish between these three functions. “Likes” have been conceived as a type of “attitude expression” (Lappas, et al., 2017) and a measure of “popularity” (Bonsón, et al., 2015); “shares” have been conceived as “advocacy” (Lappas, et al., 2017) and “participation” (Zavattaro, et al., 2015); and, “comments” have been conceived as “commitment” (Bonsón, et al., 2015), “engagement” and “networking” (Mergel, 2013b). Whereas liking and sharing a post only require individuals to click a button as a type of push button response, commenting may be considered a more involved type of interaction. Additionally, although both liking and sharing content may make content become “viral” (i.e., quickly and highly distributed across a network), people may like a post because it brings pleasure to them (Lee, et al., 2013) whereas they may share a post because they think the content is useful or interesting to others (Berger and Milkman, 2012).

To better understand the tendencies and logic of social media communication and interactions, we here review the literature on how type of content affects social media reactions. We develop hypotheses to test the effects of content or information — the communication strategies — on likes, shares, and comments separately, as the distribution seems to be heterogeneous across types of responses (Bonsón, et al., 2015). Although we recognize that “liking” is an affective reaction itself, sharing and commenting may be considered to serve non-emotional purposes (e.g., to increase knowledge, discussion, provide different perspective). Moreover, observing if “liking” behavior is indeed correlated with other emotional content (e.g., symbolic acts, marketing) provides valuable empirical evidence. However, given the limited nature of this paper and existing theory on the subject, we only develop hypotheses for some of the more prominent types of relationships expected. These hypotheses are relatively conservative and derived from previous related work.

Interaction with marketing and favorable presentation

Empirical research examining how individuals react or respond to features of marketing and favorable content on social media is limited; findings thus far suggest that the relationship is negative (Berger and Milkman, 2012; Bonsón, et al., 2015). This should not be surprising, as individuals may easily discern the nature of an advertisements or publicity campaigns as propaganda. In a study of government content on Facebook, citizens did not engage (through likes, shares, and comments) with marketing content as much as other types of content (e.g., public transportation and environment) (Bonsón, et al., 2015). In a study of German municipalities, advertisements received less comments than other types of content on Facebook. In the for-profit marketing context, the “brand name” included on posts had a significant and negative effect on the number of likes the post received (Swani, et al., 2013). “Direct calls to purchase” also had a negative effect on number of likes; however, this relationship was not significant (Swani, et al., 2013). In another for-profit context, Lee, et al. (2013) found that “informative content” (e.g., price, product availability, product mention) had a significant and negative impact on engagement (measured by likes and comments). The “informative” value, therefore, may have been interpreted by users as mere characteristics of a clear advertisement. We thus make the following hypotheses:

H1: Content coded as favorable presentation are less likely to receive likes (h1a), comments (h1b), or shares (h1c) than content not categorized as such.

H2: Content coded as marketing are less likely to receive likes (h2a), comments (h2b), or shares (h2c) than content not categorized as such.

Interaction with symbolic acts

Symbolic acts are pieces of information that reflect a social custom, politeness, or friendly formality. These are messages that refer to “phatic expressions” or “grooming talk”, and that may express, perhaps implicitly, political and social values. We assume this type of content to be more strongly related to emotional expressions (Massumi, 2015). Although such a categorical variable has not been studied in the government information context, in related research posts that were characterized with emotions, humor, and philanthropic language had a significant and positive effect on interactions (measured by number of likes and comments received) (Lee, et al., 2013). Other studies also found that the presence of emotions made content particularly viral (Stieglitz and Dang-Xuan, 2013). Content that induces “high arousal” emotions (i.e., awe, anxiety, and anger) were found to be particularly effective in making content more engaging and shared in the context of distributing news articles via e-mail (Berger and Milkman, 2012).

Emotional or affective content is not generally acknowledged in studies of government information and communication. However, in the study of Hofmann, et al. (2013), for example, non-government related information seemed to be the type of content that received the most likes, leading to the suggestion that “Facebook users are more interested in fun and private issues than in more serious reports on government activities” [3]. This type of content, however, was not theorized, and it was labeled as “other” (Hofmann, et al., 2013). Here, we find it appropriate to make the following hypothesis:

H3: Content coded as symbolic acts are more likely to receive likes (h3a), comments (h3b), or shares (h3c) than content not categorized as such.

Interaction with government information

In economics, it is understood that despite occasions in which selfish gain takes priority humans often interact with others via gift exchanges and by generating reciprocity (Fehr, et al., 1998). Therefore, online individuals may share content when there is no clear and direct gain from it but when they believe others find it useful or interesting (Berger and Milkman, 2012). However, there is little research examining how users respond to government information on social media. In a meta-analysis, “perceived usefulness” was found as one of the best predictors of intention to use e-government services (Rana, et al., 2015). Of course, in a study of “perceived usefulness”, different individuals may find different characteristics of the e-government service “useful”. Nevertheless, informative components, such as the presence of a URL for more details, has been found to be significant in predicting sharing of information (Stieglitz and Dang-Xuan, 2013). In the government context, citizens have reported expectations that government agencies share information online about its services, operations, and events (Johannessen, et al., 2012). Hofmann, et al. (2013) found that some of the most discussed/comments topics included content about “administrative news” of the government units. Given this theorization in the literature, we thus make the following hypothesis in relation to how users respond to government information:

H4: Content coded as information provision are more likely to receive likes (h4a), comments (h4b), or shares (h4c) than content not categorized as such.

Interaction with input seeking and government dialogue

It is not entirely clear from existing literature how users react and interact to instances where governments are seeking input from users or when they initiate dialogue with users via a social media platform. Based on intuitive assumptions we expect that users react positively (i.e., like, share, or further comment) to government efforts to obtain citizen feedback and provide opportunities for a citizen voice to be heard. Moreover, certain input-seeking efforts could greatly benefit from the distributive power of social networking sites. If users recognize the need for content to be shared to be effective, they may elect to share content. In the study of Johannessen, et al. (2012) citizens reported expectations that governments not only provide useful information, but also that governments engage in dialogues with the community at large and with businesses in particular. Moreover, citizens expected that governments create a forum for debates, where communicative interaction may take place (Johannessen, et al., 2012). We therefore anticipate that content that refers to input-seeking efforts, references to off-line interaction between governments and citizens, and instances where governments respond to comments from citizens, are likely to receive more virtual interaction via likes, shares, or comments. We thus hypothesize:

H5: Content coded as input seeking are more likely to receive likes (h6a), comments (h6b), or shares (h7c) than content not categorized as such.

H6: Content coded as dialogue and interaction are more likely to receive likes (h7a), comments (h7b), or shares (h7c) than content not categorized as such.

 

++++++++++

Methodology

Selection of posts and coding

To select a representative sample of local government Facebook posts to analyze, we used a stratified sampling methodology. We randomly selected four states of each major U.S. census region and then selected their largest city by population. For each of the 16 cities selected, we used the Facebook built-in search function to identify all governmental departments with a Facebook page for that city. From this set of 121 agencies, we randomly selected 55 department pages (approximately 50 percent). For each page, we retrieved all posts from 1 August 2015 to 1 August 2016 (14,482) and then randomly selected 10 percent of those posts, ultimately for a sample of 1,421 unique posts to manually code. The multiple steps of sampling in the stratification process were carried out in order to obtain a manageable amount of posts to manually code. Although we would have liked to carried out automated analyses of larger sample sizes, we believe that we obtained a relatively representative sample of government department Facebook posts for the period at question. We further justified this sampling strategy in DePaula, et al. (2018).

The posts were coded as reflecting one or more of the different types of communicated content as outlined in Table 1. Given the complexity of the content categories and multiple messages that could be inferred from a single post, our coding rules, by design, attempted to avoid double coding. However, given the nature of the message of each post it was virtually impossible to not consider double coding without disregarding important content from the post. Not counting instances of “online dialogue” (which were responses that the department gave to a user comments) eight percent of the posts were double coded. Approximately 43 percent of the double coded posts were between symbolic acts and favorable presentation, and 52 percent of the double coded posts were between operation/events with another category. Two of the authors carried out the development of coding rules during a pilot and confirmatory study. After establishing the rules, we recruited a third coder to test our categories and coding instructions. Details of the coding may be found in DePaula, et al. (2018). Ultimately, inter-coder reliability yielded almost perfect agreement (k > 0.81) for the four broader categories and almost perfect or substantial agreement (k > 0.61) for the more specific categories. All tests were significant (p < 0.001).

Analysis

The unit of analysis in this study is the Facebook post of a local government agency. The category of content is considered the independent variable, and the number of likes, comments, and shares dependent variables (the values are normalized for the number of page likes of each page, which we assume is a rough estimate of the users who have had access to the content). Normality of the data for all categories of content was tested using the Shapiro-Wilk test of normality. Homogeneity of variances was tested using Levene’s test of homogeneity. We used independent samples t-test to analyze the effect of each type of content on user response for those categories that did not violate assumptions of normality and homogeneity of variances. We followed Welch’s t-test procedure — a non-parametric alternative of t-test — in cases where the data was non-normally distributed, distributions in each group did not have the same shape, or the assumption of homogeneity of variance was violated (Siegel and Castellan, 1988).

 

++++++++++

Empirical results

In this section we provide some descriptive statistics of our sample. In Table 2 we present some statistics regarding the five most popular pages, based on percentage of city population that liked the page. All measures are for the period from August 2015 to August 2016. The table shows that a variety of government department types that are popular on Facebook, including fire, police, arts, and city departments. However, it should be noted, some of the other more popular pages are also for fire and police departments. Local government fire departments as well as police departments seem to be more often present or popular on Facebook compared to all other department types. The oldest agency on Facebook from our sample was New York’s Parks and Recreation Department, with the earliest post identified as from June 2008. As Figure 1 shows, the popularity of local government department pages seemed to follow a power law distribution: a few pages have a relatively high percentage of the city population connected on Facebook, but most government pages (i.e., 65 percent of them) have less than one percent of their respective city population ”liking“ them on Facebook. Details of the long tail may be observed on the inset graph of Figure 1. We refrained from providing aggregate measures, such as average page likes (for all pages), given the large differences of activity across all individual Facebook pages.

 

Table 2: Five most popular pages based on percentage of city population that liked the page, 2015–2016.
CityDepartmentAverage page likes
(one-year period)
Percentage increase in likes
(one-year period)
Percentage of city population who liked the pageEarliest post found
Boston, Mass.Police168,95718.826.21 February 2010
Jonesboro, Ark.Animal10,80517.915.120 July 2010
Bridgeport, Conn.Fire8,9434.36.110 December 2009
Washington D.C.Arts31,07924.94.711 February 2009
New York, N.Y.City378,15814.84.414 April 2011

 

 

Percentage of city population that likes the department page on Facebook for all 55 agencies in the sample in descending order
 
Figure 1: Percentage of city population that likes the department page on Facebook for all 55 agencies in the sample in descending order (inset graph: focus on < five percent).

 

General categories of content

In the whole sample analyzed, approximately 62 percent of posts were coded as information provision, 2.5 percent as input seeking, 8.3 percent as online dialogue/off-line interaction, and 40 percent as symbolic and presentational exchanges, as summarized in Table 3. Notice that double-coding was permissible in certain instances. A series of t-tests were carried out separately for likes, comments, and shares to compare if the use of any of the communication strategies significantly affected the quantity of user response measures (i.e., likes, comments, and shares). Independent-samples t-test was run for all general categories, given the results of normality test for these variables.

 

Table 3: General categories of content and relationship with user response measures.
Note: *p < .05
 nLikesSharesComments
Information provision885Less*LessLess*
Input seeking35LessMoreLess
Dialogue/Interaction114MoreMoreMore*
Symbolic presentation581More*MoreMore*

 

As summarized in Table 3, among the general categories, posts coded as symbolic presentation (M = 129.98, SD = 481.08) received significantly more likes than the posts that were not coded in this category (M = 24.16, SD = 80.04). Posts that reflected symbolic presentation and dialogue/interaction, as general categories, received significantly more comments than posts not coded as such. Information provision posts had significantly less likes (M = 34.33, SD = 112.10) and significantly less comments (M = 3.21, SD = 14.41) than posts that were not coded as information provision (M = 68.72, SD = 321.68; M = 7.90, SD = 23.39). The direction of relationship may be observed for other categories in Table 3, however, the differences were not significant (p < .05).

Specific categories of content

The distribution of specific categories of content and the relationship with user response is presented in Table 4. A Welch’s t-test was run for the analyses of symbolic acts and operations/events, given results from the normality tests for these variables. We used independent t-test for the other variables. As a summary of significant results, posts coded as symbolic acts (M = 145.60, SD = 338.91) and online dialogue (M = 165.40, SD = 236.49) received significantly more likes than posts that were not categorized as such (M = 46.94, SD = 308.99; and M = 62.35, SD = 320.75, respectively). Moreover, symbolic acts (M = 9.12, SD = 22.90), favorable presentation (M = 7.64, SD = 23.52), and online dialogue (M = 12.73, SD = 16.38) received significantly more comments than posts that were not categorized in these categories (M = 3.90, SD = 16.95; M = 4.41, SD = 17.14; and M = 4.58, SD = 18.47, respectively).

On the other hand, posts that were coded as operations/events information had significantly less likes (M = 37.14, SD = 120.36) and significantly less comments (M = 3.56, SD = 15.52) than the posts that were not coded in this specific category (M = 99.33, SD = 436.30; and M = 6.48, SD = 21.02). Also, posts that were coded as public service announcements had significantly less likes (M = 9.4, SD = 20.35) and significantly less comments (M = 0.45, SD = 1.63) than posts that were not coded in this category (M = 70.07, SD = 324.74 for likes; and M = 5.19, SD = 18.84 for comments).

 

Table 4: Specific categories of content and relationships with reaction measures.
Note: *p < .05
 nLikesSharesComments
Operations & events729Less*LessLess*
Public service announcement62Less*LessLess*
Feedback21LessMoreLess
Fundraising14LessLessLess
Online dialogue70More*MoreMore*
Call for discussion24LessLessLess
Call for collaboration24MoreMoreMore
Favorable presentation249MoreMoreMore*
Political positioning27MoreMoreMore
Symbolic act295More*LessMore*
Marketing67MoreMoreMore

 

 

++++++++++

Discussion: The affective bias

This study analyzed the relationship between user response and the types of social media content created by local government departments in the U.S. for Facebook. We first hypothesized that posts of information provision would be positively related with measures of likes, shares, and comments. This was based on the notion that individuals may appreciate useful information about their local governments and would demonstrate this appreciation by engaging with content, through liking, sharing and/or commenting. Since the analysis assumes those who engage with posts already need to “follow” an organization, this was not a particularly difficult test. We found that content coded as policy, operations and events received significantly less likes and comments than content coded in other categories. The same was the case for posts coded as public service announcements.

Posts coded as favorable presentation, which we hypothesized to affect the message negatively, received significantly more comments that posts coded in other categories. Although not statistically significant, posts coded as favorable presentation and posts coded as marketing, which we also assumed would affect user response negatively, actually received more likes and more shares than posts not coded as such. We hypothesized that posts coded as input seeking would receive more likes, shares, or comments. However, these differences were not significant and posts of this general category received on average less likes, shares, and comments than posts coded in other categories. As predicted, posts coded as symbolic acts received significantly more likes and comments than posts that did not receive this designation, and posts coded as online dialogue also received significantly more likes and comments that posts not coded in this designation. However, it should be noted, “online dialogue” refers to posts that had a government response to a user comment. These posts are more likely to have more comments that other types of posts. However, this fact itself does not explain why posts that had government responses received significantly more likes than posts that did not receive this designation.

We interpret these results as a reflection of the affective and image-based tendency of social media behavior and a lack of manifested interest for content that is simply of a factual nature or linguistically elaborate, what we may term an affective bias. At the turn of the new millennium, researchers in neuropsychology (e.g., Damasio, 1994), humanities (e.g., Sedgwick, 2003), and politics (e.g., Massumi, 2015) began to develop a notion of “affect” in their interpretation of how human experiences are characterized in the twenty-first century. It is not through ideology, but through affect that “the public is mobilized to act” [4] politically. Our experiences are mediatic and mediated (e.g., through mobile computers, Internet, interfaces), what Grusin refers to as “mediality”. Although there are many interpretations of how affect and mediality work in the world today (e.g., Leys, 2011), these concepts seem, to a lesser or greater extent, necessary to understand patterns of behavior on social media sites. As Grusin (2010) writes of social media:

interacting with such sites is made pleasurable or desirable in part because they work to produce and maintain positive affective relations with their users, to set up affective feedback loops that make one want to proliferate one’s media transactions. Indeed, something as seemingly innocuous as the fact that Facebook offers its users the option to “like” or “unlike” an item but not to “dislike” it epitomizes its bias towards fostering positive individual and collective affect. [5]

Facebook still does not offer the option to “dislike”, however, it now includes more emotion buttons (i.e., “love”, “haha”, “wow”, “sad”, “angry”) to be used as recorded reactions to posts. Although this shows that the bias of these systems is not only to foster “positive” affect (e.g., you can now post an “angry” face), the site continues to epitomize the “affective intensity” of these platforms (Paasonen, 2016). Grusin (2010) writes that affect is a type of emotional or subjective feeling “prior to and independent of [the] cognitive impact or interpretation” [6] that an event produces. As such, affectivity is distinguished and contrasted with ideology, which are rationalized and, at least to some extent, reasoned systems of belief. Although ideology may still be in place, its materialization is through affect and the affect itself becomes more important than the ideology (Massumi, 2015). In the political domain, power itself no longer operates through normative ideas based on ideological explanations but through affect itself. As Brian Massumi (2015) puts it in an interview:

The legitimisation of political power, of state power, no longer goes through the reason of state and the correct application of governmental judgment. It goes through affective channels. ... Affect is now much more important for understanding power, even state power narrowly defined, than concepts like ideology.

Grusin further suggests that there is a strong relation between how affect operates on the world today and how “mediality” is experienced in everyday life. Mediality contrasts with a notion of “representationality” — that is, representational truth. Epitomized by social network sites but affecting all mediated interaction, mediality is more concerned with “mobilizing people or populations” [7], than with any type of factual and correct reference to some external, objective phenomenon in the world. Given this understanding it becomes possible to explain the empirical results obtained in this study. From Table 4, one should note that the first six categories of communication, all referring to factual and operational information provision and efforts to obtain information and feedback from citizens, correspond, generally, to less forms of social media interaction. This is the case for 11 of the 12 categories under information provision and input seeking studied (i.e., only efforts for citizen feedback were associated with more shares but not significantly). In contrast, 11 of the 12 categories under symbolic presentation were associated with more interaction in general (i.e., only symbolic acts were associated with less shares but not significantly).

The consistency in the direction of user interaction with our broad categories (e.g., less for information provision in general; more for symbolic presentation in general) gives some support that the categories form a distinct entity. At the same time, our results also suggest that there is homogeneity on how users react via likes, shares, and comments. Studies on this type of comparison have not found consistent results (e.g., Bonsón, et al., 2015; Lee, et al., 2013). Nevertheless, the consistency of direction in our results for the three types of responses are explained by the theory that the underlying mechanisms of generating interaction on social network sites is one of affective intensity. Therefore, responses of “likes”, “shares” and “comments” are all going to be driven by sentimentality rather than the possibility for debate or policy discussion.

The presence of online dialogue by the government agency received significantly more likes and comments compared to all posts without dialogue. Once governments engage, citizens further engage and respond. What needs to be further examined are the characteristics of this dialogue and their actual adoption as policy. Research on government social media communication often assumes that if engagement is happening, it is positive and commendable. In this work, we have observed that responses from the government agency were appreciated by the public (i.e., via likes). However, we did not investigate the quality of conversation on the site. Recent evidence from China (Medaglia and Zhu, 2016) suggests that social network sites (in this case Weibo) are not sites of deliberation as exemplified by the ideal public sphere of Habermas (1989) or any prerogative for informed and reasoned discussion. Indeed, we may explain the greater levels of user interaction to government comments precisely based on the emotional nature of social media responses. Personalized attention may affect reaction behavior. However, we can also conceivably propose that reaction behavior is a sign of appreciation for the receipt of informative and therefore valuable content. We leave these questions for further research.

 

++++++++++

Conclusions and future work

Our empirical study found evidence that user interaction or reaction behavior on Facebook pages of local governments in the U.S. are significantly more often associated with symbolic and image-based content than content not characterized as such. At the same time, factual policy information and information about operations and events are associated with less likes, shares, and comments in general, and statistically significantly for some categories. This is explained by the affective nature of social media behavior which does not emphasize critical and reasoned dialogue but pre-cognitive and emotional behavior. Although studies on government social media have similar findings on how users respond to governmental content, they have suggested governments should follow these tendencies and further deploy content that will attract likes, shares, and comments (e.g., Mergel, 2013a; Zavattaro, et al., 2015). However, we question whether these government information strategies would assist with a politically informed dialogue. Should we characterize government publication of viral pictures and one-sided self-presentations as measures of transparency? Nevertheless, we found variations in how departments adopt distinct information strategies, which needs to be more systematically analyzed. Moreover, there were positive measures of interaction in terms of government dialogue, which may point to the value of government information to citizens via social media. Ultimately, compared to Web sites, the platforms provide a more virtual interactive environment through which governments may communicate with citizens. Also, there may be a rationale for the principle of affective intensity, as emotions, of course, are part of human interaction. However, we also expect governments are going to use platforms to create an unchallenged image of themselves and that the tendency of social media reactions are toward positive imagery and high-arousal. We thus suggest that the principles of open and democratic government, including the goals of transparency and public deliberation, will not be so easily realized on social media. End of article

 

About the authors

Nic DePaula is a Ph.D. candidate in the Department of Information Science and lecturer in the Departments of Information Technology Management and Information Science at the University at Albany, State University of New York. He conducts research in the areas of information policy, political communication and media studies.
E-mail: ndepaula [at] albany [dot] edu

Ersin Dincelli is a Ph.D. candidate in the Department of Information Science and an adjunct professor in the Department of Information Technology Management and Information Science at the University at Albany, the State University of New York. His primary research interests include individual behavior in the context of information security and privacy, cybersecurity education, and cross-cultural issues in information systems.

 

Notes

1. Hofmann, et al., 2013, p. 391.

2. Smail, 2008, p. 176.

3. Hofmann, et al., 2013, p. 391.

4. Grusin, 2010, p. 77.

5. Grusin, 2010, p. 3.

6. Grusin, 2010, p. 79.

7. Ibid.

 

References

Peter Bellström, Monika Magnusson, John Sören Pettersson, and Claes Thorén, 2016. “Facebook usage in a local government A content analysis of page owner posts and user posts,” Transforming Government: People, Process and Policy, volume 10, number 4, pp. 548–567.
doi: https://doi.org/10.1108/TG-12-2015-0061, accessed 17 March 2018.

Jonah Berger and Katherine L. Milkman, 2012. “What makes online content viral?” Journal of Marketing Research, volume 49, number 2, pp. 192–205.
doi: https://doi.org/10.1108/TG-12-2015-0061, accessed 17 March 2018.

John C. Bertot, Paul T. Jaeger, and Justin M. Grimes, 2010. “Using ICTs to create a culture of transparency: E-government and social media as openness and anti-corruption tools for societies,” Government Information Quarterly, volume 27, number 3, pp. 264–271.
doi: https://doi.org/10.1108/TG-12-2015-0061, accessed 17 March 2018.

Enrique Bonsón, Sonia Royo, and Melinda Ratkai, 2015. “Citizens’ engagement on local governments’ Facebook sites. An empirical analysis: The impact of different media and content types in Western Europe,” Government Information Quarterly, volume 32, number 1, pp. 52–62.
doi: https://doi.org/10.1016/j.giq.2014.11.001, accessed 17 March 2018.

Tony Bovaird, 2007. “Beyond engagement and participation: User and community coproduction of public services,” Public Administration Review, volume 67, number 5, pp. 846–860.
doi: https://doi.org/10.1111/j.1540-6210.2007.00773.x, accessed 17 March 2018.

danah m. boyd and Nicole B. Ellison, 2007. “Social network sites: Definition, history, and scholarship,” Journal of Computer-Mediated Communication, volume 13, number 1, pp. 210–230.
doi: https://doi.org/10.1111/j.1083-6101.2007.00393.x, accessed 17 March 2018.

Gordon C. Bruner, 1990. “Music, mood, and marketing,” Journal of Marketing, volume 54, number 4, pp. 94–104.
doi: http://dx.doi.org/10.2307/1251762, accessed 17 March 2018.

Antonio Damasio, 1994. Descartes’ error: Emotion, reason, and the human brain. New York: Putnam.

Nic DePaula and Ersin Dincelli, 2016. “An empirical analysis of local government social media communication: Models of e-government interactivity and public relations,” dg.o ’16: Proceedings of the 17th International Digital Government Research Conference on Digital Government Research, pp. 348–356.
doi: http://dx.doi.org/10.1145/2912160.2912174, accessed 17 March 2018.

Nic DePaula, Ersin Dincelli, and Teresa M. Harrison, 2018. “Toward a typology of government social media communication: Democratic goals, symbolic acts and self-presentation,” Government Information Quarterly, volume 35, number 1, pp. 98–108.
doi: https://doi.org/10.1016/j.giq.2017.10.003, accessed 17 March 2018.

José van Dijck, 2013. The culture of connectivity: A critical history of social media. New York: Oxford University Press.

Judith Donath, 2007. “Signals in social supernets,” Journal of Computer-Mediated Communication, volume 13, number 1, pp. 231–251.
doi: https://doi.org/10.1111/j.1083-6101.2007.00394.x, accessed 17 March 2018.

Robin Dunbar, 1996. Grooming, gossip, and the evolution of language. Cambridge, Mass.: Harvard University Press.

Ernst Fehr, Georg Kirchsteiger, and Arno Riedlc, 1998. “Gift exchange and reciprocity in competitive experimental markets,” European Economic Review, volume 42, number 1, pp. 1–34.
doi: https://doi.org/10.1016/S0014-2921(96)00051-7, accessed 17 March 2018.

Christian Fuchs, 2017. Social media: A critical introduction. Thousand Oaks, Calif.: Sage.

Jennifer Golbeck, Justin M. Grimes, and Anthony Rogers, 2010. “Twitter use by the U.S. Congress,” Journal of the American Society for Information Science and Technology, volume 61, number 8, pp. 1,612–1,621.
doi: https://doi.org/10.1002/asi.21344, accessed 17 March 2018.

Melissa W. Graham, Elizabeth J. Avery, and Sejin Park, 2015. “The role of social media in local government crisis communications,” Public Relations Review, volume 41, number 3, pp. 386–394.
doi: https://doi.org/10.1016/j.pubrev.2015.02.001, accessed 17 March 2018.

Shannon Greenwood, Andrew Perrin, and Maeve Duggan, 2016. “Social media update 2016,” Pew Research Center (11 November), at http://www.pewinternet.org/2016/11/11/social-media-update-2016/, accessed 30 November 2016.

Richard Grusin, 2010. Premediation: Affect and mediality after 9/11. New York: Palgrave Macmillan.

Jürgen Habermas, 1989. The structural transformation of the public sphere: An inquiry into a category of bourgeois society. Translated by Thomas Burger with the assistance of Frederick Lawrence. Cambridge, Mass.: MIT Press.

Loni Hagen, Thomas Keller, Stephen Neely, Nic DePaula, and Claudia Robert-Cooperman, 2017. “Crisis communications in the age of social media: A network analysis of Zika-related tweets,” Social Science Computer Review (21 August).
doi: https://doi.org/10.1177/0894439317721985, accessed 17 March 2018.

Teresa M. Harrison, Santiago Guerrero, G. Brian Burke, Meghan Cook, Anthony Cresswell, Natalie Helbig, Jana Hrdinova, and Theresa Pardo, 2012. “Open government and e-government: Democratic challenges from a public value perspective,” Information Polity, volume 17, number 2, pp. 83–97.
doi: https://doi.org/10.3233/IP-2012-0269, accessed 17 March 2018.

Sara Hofmann, Daniel Beverungen, Michael Räckers, and Jörg Becker, 2013. “What makes local governments’ online communications successful? Insights from a multi-method analysis of Facebook,” Government Information Quarterly, volume 30, number 4, pp. 387–396.
doi: https://doi.org/10.3233/IP-2012-0269, accessed 17 March 2018.

Marius Rohde Johannessen, Leif Skiftenes Flak, and Øystein Saebø, 2012. “Choosing the right medium for municipal eparticipation based on stakeholder expectations,” In: Efthimios Tambouris, Ann Macintosh, and Øystein Saebø (editors). Electronic participation. Lecture Notes in Computer Science, volume 7444. Berlin: Springer, pp. 25–36.
doi: https://doi.org/10.1007/978-3-642-33250-0_3, accessed 17 March 2018.

Andrea L.Kavanaugh, Edward A.Fox, Steven D.Sheetz, Seungwon Yang, Lin Tzy Li, Donald J.Shoemaker, Apostol Natsev, and Lexing Xie, 2012. “Social media use by government: From the routine to the critical,” Government Information Quarterly, volume 29, number 4, pp. 480–491.
doi: https://doi.org/10.1016/j.giq.2012.06.002, accessed 17 March 2018.

Mathias Klang and Jan Nolin, 2011. “Disciplining social media: An analysis of social media policies in 26 Swedish municipalities,” First Monday, volume 16, number 8, at http://firstmonday.org/article/view/3490/3027, accessed 18 October 2017.
doi: http://dx.doi.org/10.5210/fm.v16i8.3490, accessed 17 March 2018.

G. Lappas, A. Triantafillidou, A. Deligiaouri, and A. Kleftodimos, 2017. “Facebook communication strategies and their effectiveness: A case study of Greek local municipal governments,” MISNC ’17: Proceedings of the 4th Multidisciplinary International Social Networks Conference, article number 22.
doi: http://dx.doi.org/10.1145/3092090.3092114, accessed 17 March 2018.

Dokyun Lee, Kartik Hosanagar, and Harikesh Nair, 2013. “Advertising content and consumer engagement on social media: Evidence from Facebook,” Management Science, at https://ssrn.com/abstract=2290802, accessed 3 December 2016.

Ruth Leys, 2011. “The turn to affect: A critique,” Critical Inquiry, volume 37, number 3, pp. 434–472.
doi: http://dx.doi.org/10.1086/659353, accessed 17 March 2018.

Dennis Linders, 2012. “From e-government to we-government: Defining a typology for citizen coproduction in the age of social media,” Government Information Quarterly, volume 29, number 4, pp. 446–454.
doi: https://doi.org/10.1016/j.giq.2012.06.003, accessed 17 March 2018.

Brian Massumi, 2015. Politics of affect. Cambridge: Polity.

Rony Medaglia and Demi Zhu, 2016. “Paradoxes of deliberative interactions on government-managed social media: Evidence from China,” dg.o ’16: Proceedings of the 17th International Digital Government Research Conference on Digital Government Research, pp. 435–444.
doi: https://doi.org/10.1145/2912160.2912184, accessed 17 March 2018.

Albert Meijer and Marcel Thaens, 2013. “Social media strategies: Understanding the differences between North American police departments,” Government Information Quarterly, volume 30, number 4, pp. 343–350.
doi: https://doi.org/10.1016/j.giq.2013.05.023, accessed 17 March 2018.

Ines Mergel, 2016. “Social media institutionalization in the U.S. federal government,” Government Information Quarterly, volume 33, number 1, pp. 142–148.
doi: https://doi.org/10.1016/j.giq.2015.09.002, accessed 17 March 2018.

Ines Mergel, 2013a. “A framework for interpreting social media interactions in the public sector,” Government Information Quarterly, volume 30, number 4, pp. 327–334.
doi: https://doi.org/10.1016/j.giq.2013.05.015, accessed 17 March 2018.

Ines Mergel, 2013b. “Social media adoption and resulting tactics in the U.S. federal government,” Government Information Quarterly, volume 30, number 2, pp. 123–130.
doi: https://doi.org/10.1016/j.giq.2012.12.004, accessed 17 March 2018.

Karen Mossberger, Yonghong Wu, and Jared Crawford, 2013. “Connecting citizens and local governments? Social media and interactivity in major U.S. cities,” Government Information Quarterly, volume 30, number 4, pp. 351–358.
doi: https://doi.org/10.1016/j.giq.2013.05.016, accessed 17 March 2018.

Ashwini Nadkarni and Stefan G. Hofmann, 2012. “Why do people use Facebook?” Personality and Individual Differences, volume 52, number 3, pp. 243–249.
doi: https://doi.org/10.1016/j.paid.2011.11.007, accessed 17 March 2018.

Barack Obama, 2009. “Transparency and open government: Memorandum for the heads of executive departments and agencies,” White House (21 January), at https://obamawhitehouse.archives.gov/the-press-office/transparency-and-open-government, accessed 17 March 2018.

Susanna Paasonen, 2016. “Fickle focus: Distraction, affect and the production of value in social media,” First Monday, volume 21, number 10, at http://firstmonday.org/article/view/6949/5629, accessed 16 October 2017.
doi: http://dx.doi.org/10.5210/fm.v21i10.6949, accessed 17 March 2018.

Zizi Papacharissi and Maria de Fatima Oliveira, 2012. “Affective news and networked publics: The rhythms of news storytelling on #Egypt,“ Journal of Communication, volume 62, number 2, pp. 266–282.
doi: https://doi.org/10.1111/j.1460-2466.2012.01630.x, accessed 17 March 2018.

Joel Penney, 2015. “Social media and symbolic action: Exploring participation in the Facebook red equal sign profile picture campaign,“ Journal of Computer-Mediated Communication, volume 20, number 1, pp. 52–66.
doi: https://doi.org/10.1111/jcc4.12092, accessed 17 March 2018.

Nripendra P. Rana, Yogesh K. Dwivedi, and Michael D. Williams, 2015. ”A meta-analysis of existing research on citizen adoption of e-government,“ Information Systems Frontiers, volume 17, number 3, pp. 547–563.
doi: https://doi.org/10.1007/s10796-013-9431-z, accessed 17 March 2018.

Rodrigo Sandoval-Almazan and J Ramon Gil-Garcia, 2012. ”Are government Internet portals evolving towards more interaction, participation, and collaboration? Revisiting the rhetoric of e-government among municipalities,“ Government Information Quarterly, volume 29, supplement 1, pp. s72–s81.
doi: https://doi.org/10.1016/j.giq.2011.09.004, accessed 17 March 2018.

Rodrigo Sandoval-Almazan, David Valle-Cruz, and Andrea L. Kavanaugh, 2018. “The diffusion of social media among state governments in Mexico,” International Journal of Public Administration in the Digital Age, volume 5, number 1, pp. 63–81.
doi: https://doi.org/10.4018/IJPADA.2018010104, accessed 17 March 2018.

John R. Searle, 1969. Speech acts: An essay in the philosophy of language. London: Cambridge University Press.

Eve Kosofsky Sedgwick, 2003. Touching feeling: Affect, pedagogy, performativity. Durham, N.C.: Duke University Press.

Pamela J. Shoemaker (editor), 1989. Communication campaigns about drugs: Government, media, and the public. Hillsdale, N.J.: L. Erlbaum Associates.

Sidney Siegel and N. John Castellan, 1988. Nonparametric statistics for the behavioral sciences. Second edition. New York: McGraw-Hill.

Daniel Lord Smail, 2008. On deep history and the brain. Berkeley: University of California Press.

Stefan Stieglitz and Linh Dang-Xuan, 2013. “Emotions and information diffusion in social media — Sentiment of microblogs and sharing behavior,” Journal of Management Information Systems, volume 29, number 4, pp. 217–248.
doi: https://doi.org/10.2753/MIS0742-1222290408, accessed 17 March 2018.

Maddle Stone, 2015. “8 government Instagram accounts you should definitely follow,” Gizmodo (22 July), at https://gizmodo.com/8-government-instagram-accounts-you-should-definitely-f-1719387212, accessed 12 January 2018.

Kunal Swani, George Milne, and Brian P. Brown, 2013. “Spreading the word through likes on Facebook: Evaluating the message strategy effectiveness of Fortune 500 companies,” Journal of Research in Interactive Marketing, volume 7, number 4, pp. 269–294.
doi: https://doi.org/10.1108/JRIM-05-2013-0026, accessed 17 March 2018.

Sharon Tettegah (editor), 2016. Emotions, technology, and social media. London: Academic Press.

Jeffrey W. Treem and Paul M. Leonardi, 2012. “Social media use in organizations: Exploring the affordances of visibility, editability, persistence, and association,” Annals of the International Communication Association, volume 36, number 1, pp. 143–189.
doi: https://doi.org/10.1080/23808985.2013.11679130, accessed 17 March 2018.

Zeynep Tufekci, 2008. “Grooming, gossip, Facebook and MySpace: What can we learn about these sites from those who won’t assimilate?” Information, Communication & Society, volume 11, number 4, pp. 544–564.
doi: https://doi.org/10.1080/13691180801999050, accessed 17 March 2018.

Richard D. Waters and Jensen M. Williams, 2011. “Squawking, tweeting, cooing, and hooting: Analyzing the communication patterns of government agencies on Twitter,” Journal of Public Affairs, volume 11, number 4, pp. 353–363.
doi: https://doi.org/10.1002/pa.385, accessed 17 March 2018.

Staci M. Zavattaro, P. Edward French, and Somya D. Mohanty, 2015. “A sentiment analysis of U.S. local government tweets: The connection between tone and citizen involvement,” Government Information Quarterly, volume 32, number 3, pp. 333–341.
doi: https://doi.org/10.1016/j.giq.2015.03.003, accessed 17 March 2018.

 


Editorial history

Received 2 March 2018; accepted 14 March 2018.


Creative Commons License
This paper is licensed under a Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International License.

Information strategies and affective reactions: How citizens interact with government social media content
by Nic DePaula and Ersin Dincelli.
First Monday, Volume 23, Number 4 - 2 April 2018
https://journals.uic.edu/ojs/index.php/fm/article/view/8414/6695
doi: http://dx.doi.org/10.5210/fm.v23i4.8414





A Great Cities Initiative of the University of Illinois at Chicago University Library.

© First Monday, 1995-2019. ISSN 1396-0466.