Do platforms favour dissidents? Characterizing political actor types based on social media uses and gratifications

Abstract

BACKGROUND:

The rise of social media has resulted in a dramatic change in citizen engagement in political processes. This raises the question of whether affordances of social network sites motivate alternative politics more than more conventional form of political engagement.

OBJECTIVE:

1) identify differences in social media uses and gratifications among four political personality types (i.e., potential dissidents, allegiants, subordinates, and the alienated), and 2) examine the extent to which political personality types can be discerned using social media uses and gratifications.

METHODS:

313 United States citizens above the age of 18 completed a survey using the revised MAIN model scale to measure social media uses and gratifications. Subjects were categorised into political personality types based on the Gamson Hypothesis and Paige’s conceptualisation of actor types. We developed a multinomial logistic regression model to examine the relationship between predictors (uses and gratifications) and political personality types.

RESULTS:

Potential allegiants and dissidents are driven by a similar set of social media uses and gratifications as opposed to political subordinates and the alienated.

CONCLUSION:

Social media can provide more gratifications for potential dissidents and allegiants, ‘favouring’ personality types with high political efficacy.

Keywords

Social media uses and gratifications uses and grats 2.0 affordances dissidents allegiant

Dr. Chamil Rathnayake is a Lecturer at Strathclyde University. He takes an interdisciplinary approach to understand new media platforms. His current research pays special attention to ‘mediated publicness’ enabled by social media affordances. He is a computational social scientist, applying methods in Social Network Analysis (SNA) and Natural Language Processing, such as centrality analysis, community detection, text mining, clustering, and topic model analysis. He completed his doctoral studies at the University of Hawaii-Manoa, focusing on social informatics, communication theory, and communication policy and planning.

Dr. Jenifer Sunrise Winter is a Professor in the School of Communications and Co-Director of the Pacific Information and Communication Technology for Development Collaborative (PICTDC) at the University of Hawaii at Manoa. Her research addresses data governance, policy, and ethics related to big data, artificial intelligence (AI), and the Internet of Things (IoT). She is Associate Editor of the Journal of Information, Communication and Ethics in Society and Co-Chair of the Research Committee at the Pacific Telecommunications Council. Dr. Winter is chair of the interdisciplinary Communication and Information Sciences PhD program.

1 Introduction

The rise of social media has resulted in a dramatic change in citizen engagement in political processes. This change can be largely attributed to the fact that social media, as Skoric, Zhu, Goh, and Pang [1] note, offer informational, expressive, and relational uses that can motivate engagement. The current social media discourse includes many studies that examine online political engagement from the perspective of activism and dissidence (e.g., [2 –13]). This raises the question of whether affordances of social network sites motivate alternative politics more than more conventional forms of political engagement. Although previous studies highlight the ‘revolutionary potential’ of the Internet, several scholars, those who examine online political polarization and echo chambers in particular (e.g., [14]), indicate that the presence of opposing political perspectives on social network sites is not negligible. Moreover, there is a growing number of research studies that explore online political activity from different points of view, such as Engesser, Ernst, Esser, and Büchel’s [15] study that demonstrates how populist politics manifest on social media, Pearce’s [16] work on the subtle use of social media for authoritarian control, and Groshek and Koc-Michalska’s [17] research on social media support for both Democratic and Republican populist political figures.

Recent research shows that online engagement depends on a range of user-related factors. For instance, Russo and Amnå [18] argue that extraverts and those who are open to new experiences tend to participate in political activity more than conscientious individuals. Bimber, Cunill, Copeland, and Gibson’s [19] study shows moderating effects of political interest on the relationship between digital media use and political participation. Moreover, Boulianne’s [20] work identifies direct and indirect effects of online news consumption on civic and political engagement. While these studies shed light on the topic, there is a dearth of work that connects general social media use, especially uses and gratifications, with political differences. Motivations for social media use span beyond political purposes. While some social media gratifications, such as affection [21], may appeal to those who do not seek to engage in online politics, other gratifications, such as obtaining information and socialising [22], may encourage political engagement. Therefore, differences in perceived uses and gratifications may, at least to some extent, help distinguish between those who have different political attributes. Accordingly, work that examines whether social media serve as venues that facilitate political dissidence should take into account gratifications that motivate such tendencies as well as other types of political behaviour (e.g., allegiance and subordination). This study examines whether social media uses and gratifications can accurately reveal political differences. This can be examined on two different levels. On one level, researchers can examine if social media data can reveal direct attributes, such as the political orientation or party affiliation of users. On a more abstract level, studies can focus on whether social media consumption can help classify users into categories that display distinct political actor types. This study is concerned with the latter. Accordingly, we test differences in social media uses and gratifications between four political actor types (dissidents, allegiants, subordinates, and the alienated). Examining political differences from the perspective of general social media use is important for several reasons. From a theoretical viewpoint, it shows the relevance of constructs which remain beyond the study of political motivations in examining political behaviour. From a policy perspective, such work can help design social media platforms to encourage political engagement among those who have different political attributes.

1.1 Related literature

We approach the study of social media uses and gratifications focusing on differences among potential political dissidents, allegiants, subordinates, and the alienated. According to Paige [23], combinations of political trust and efficacy display differences among political actor types (allegiants- high efficacy, high trust; subordinates- high trust, low efficacy; dissidents- high efficacy, low trust, and alienated- low trust, low efficacy). This approach is different from the study of conventional political actors, “individuals who have obtained at least some measure of political power and/or authority in a particular society who engage in activities that can have a significant influence on decisions, policies, media coverage, and outcomes associated with a given conflict” [24, p.1]. Our approach shifts the attention from conventional political actors, such as elected officials and policy makers, to different groups within the general public that display different behavioural patterns. Such work is important for several reasons. First, although social media studies examine online dissidence in a range of contexts (e.g.,[12 , 25–27]), the field lacks academic work on the ways in which social media gratify those who have ‘dissident potential’. The need to examine connections between potential dissidence — representing a layer of social media users who may transform into active dissidents in democratic contexts—and new media use relates to the fact that the accessibility and ease of use of those platforms may encourage less active individuals to engage in more active forms of democratic dissidence.

Second, there is a need for studies that focus on political allegiance on social network sites. Social media may provide a range of uses and gratifications for political allegiants –those who trust the existing political system–because, as Rathnayake and Winter [28] argue, social media gratify users with different political attributes. Early work done on mobilization potential [29], as well as subsequent studies [30, 31], provide a theoretical background to capture potential dissidence and allegiance in the context of social media. Third, political subordinates and the alienated remain largely unexplored in social media studies.

Online platforms offer a multitude of benefits for users. For instance, Facebook, ‘the pleasure machine’, as Vaidhyanathan [32] calls it, enables people to derive value from reconnecting, interaction, joining causes, and entertainment. These perceived benefits may help differentiate between different types of users. Uses and Gratifications (U&G), a theoretical approach originating in media studies [33] that captures the essence of benefits perceived by users, views media consumption as a function of the needs it gratifies. Scholars have used this approach to examine social media consumption [34 –37]. The U&G approach serves as a foundation for discerning political differences, as it helps understand underlying motivations behind media consumption. Recent work suggests that the field of social media uses and gratifications needs academic work that captures nuances in new media platforms. For instance, Sundar and Limperos [38] show biases in previous literature towards social and psychological factors. This exposes a rich field for potential inquiry, as the fact that social media are a new domain defined mainly based on user-generated content demands more platform-oriented studies that consider affordances of new media. Sundar and Limperos provide a conceptual foundation for such a line of inquiry by proposing an affordance-driven framework, the MAIN model, which captures the ability of social media to create new gratifications. Rathnayake and Winter [39] followed this work by validating their Uses and Grats 2.0 framework.

1.2 Social media uses and gratifications

The U&G perspective, a theoretical approach that dates back to the origins of mass communication research [40], holds that media consumption is intrinsically connected with the need for gratification. According to Haridakis [41, p.2], the U&G approach suggests that “people’s social and psychological characteristics influence their needs and desires which are manifested in their motives for using communication channels.” U&G studies include a wide range of research that has evolved with the transformations in media and communication technology. According to Papacharissi [42], a typical U&G study can look at a specific medium, compare it with another medium, and study different aspects, such as motives, social and psychological antecedents, and effects of media consumption. Starting from work that stressed the need for U&G studies in the context of the Internet [33], social media U&G research has expanded rapidly during the past decade, focusing on different platforms and effects of a wide range of variables [e.g., 22, 28, 34, 37, 43–48]. Previous studies have also highlighted different uses and gratifications perceived by social media users, such as information seeking, entertainment, and social utility [31], social interaction, and information seeking [49], and socializing and self-status seeking [22]. There is a growing amount of research examining social media uses and gratifications in the context of politics. Ancu and Cozma’s [49] work, for example, identifies interaction as the primary motive for accessing candidate profiles on social network sites. Park, Kee, and Valenzuela [22] claim that informational uses of social networks are connected to civic and political action, while entertainment does not explain the extent of political participation.

The notion of affordances—the relational structure between user and technology that allows or constrains behaviour within a given context [50]—has permeated research that examines user activity on social network sites. Coined by Gibson [51], this concept suggests that affordances of a given environment should be understood relative to their context. Within the context of social media, new affordances allow a range of new behaviours that were not possible in the conventional media setting. Sundar and Limperos [38] stress the need for conceptualizing uses and gratifications based on new media affordances and argue that many studies have only slightly modified older media gratifications to suit new media. They suggest an affordance-based framework (the MAIN model) that includes 16 gratifications representing four broad classes of affordances (Modality, Agency, Interactivity, and Navigability). Sundar [52] argues that each user is different in terms of the meanings he or she makes from affordances. The MAIN Model suggests that modality cues, agency cues, interactivity cues, and navigability cues embedded in affordances are significant in shaping user assessment of the medium. Sundar argues that the sheer presence of these affordances results in users experiencing media content in a certain way. In general, the MAIN Model rejects the idea that all gratifications relate to innate needs and argues that new and distinctive gratifications can emerge from new media affordances. The model emphasizes the capability of a medium to facilitate certain actions and suggests that the user is an integral element as he or she interprets the affordance. The MAIN model suggested a pool of items that led to the development of an affordance-based scale for social media U&G, which was converted into a social media uses and gratifications scale by Rathnayake and Winter [39]. Validation of the scale led to a revision to Sundar and Limperos’ conceptual framework, as results suggested combining several constructs (e.g., Coolness and Novelty, and Browsing and Scaffolding/Navigation Aids). The revised MAIN framework for social media U&G is presented in Table 1.

Table 1
The Revised MAIN Framework

MAIN Dimension Definitions of Gratifications

Modality: The presence of multiple modes of information Realism: The ability to access media content that can resemble real-life situations and contexts

Coolness: Positive perceptions that indicate appreciation of style, newness, and attractiveness of something in social contexts.

Being There: The ability of social media platforms to create a sense of being present in a given environment

Agency: Users being sources of information Agency-enhancement: Gratification of being a source of content, rather than a passive recipient of information

Community-building: The ability to connect with other social media users and sustain short/long term social networks for various purposes

Filtering/Tailoring: The ability to control information shared by others on a user’s social media pages and control over information shared by the user

Interactivity: Ability to interact with the platform Activity: Active engagement of the user in the media use process, via frequent use of input mechanisms

Responsiveness: Perceived ability of media platforms to respond to user commands/input

Navigability: Ability to navigate through different types of content and enjoy the platform use process Browsing/Variety-seeking: Perceived diversity of media content by using design features, such as links and menus.

Play/Fun: Perceived enjoyment through the media use process.

MAIN Dimension	Definitions of Gratifications
Modality: The presence of multiple modes of information	Realism: The ability to access media content that can resemble real-life situations and contexts
	Coolness: Positive perceptions that indicate appreciation of style, newness, and attractiveness of something in social contexts.
	Being There: The ability of social media platforms to create a sense of being present in a given environment
Agency: Users being sources of information	Agency-enhancement: Gratification of being a source of content, rather than a passive recipient of information
	Community-building: The ability to connect with other social media users and sustain short/long term social networks for various purposes
	Filtering/Tailoring: The ability to control information shared by others on a user’s social media pages and control over information shared by the user
Interactivity: Ability to interact with the platform	Activity: Active engagement of the user in the media use process, via frequent use of input mechanisms
	Responsiveness: Perceived ability of media platforms to respond to user commands/input
Navigability: Ability to navigate through different types of content and enjoy the platform use process	Browsing/Variety-seeking: Perceived diversity of media content by using design features, such as links and menus.
	Play/Fun: Perceived enjoyment through the media use process.

Source: Modified from Rathnayake and Winter (2017).

Modality, defined by Sundar and Limperos [38] as “different methods of presentation (e.g., audio or pictures) of media content, appealing to different aspects of the human perceptual system (e.g., hearing, seeing)” (p.512), acknowledges that the Internet can provide users with content in multiple modalities and some of them can be considered unique (e.g., pop-up ads). The modality affordance results in three gratifications (Realism, Coolness, and Being There).The agency affordance of the MAIN Model suggests that the Internet allows users to be agents or sources of information, and this view acknowledges the ability of users to be gatekeepers of content, build communities, and contribute. Agency affordance can lead to Agency-enhancement, Community Building, Bandwagon, and Filtering gratifications. Interactivity, according to Sundar [52], is the most distinctive affordance of digital media and it relates to real-time interaction and activity on a given medium. The revised interactivity affordance [39] includes two gratifications (Activity and Responsiveness). Finally, Sundar and Limperos [38] define navigability as “the affordance that allows user movement through the medium” (p.516). Navigability can include the expectation that users are scaffolded through certain processes (e.g., online transactions) and will encounter some fun elements. Accordingly, the navigability affordance can result in two gratifications (Browsing and Play). Table 1 defines each affordance.

1.3 Political actor types

A large number of studies stress the role social media play in explicit political dissent [e.g., 7, 53]. Although social media can gratify users who have different political attributes [28], previous studies pay primary attention to political dissidence. As we argued before, individuals with different political orientations, such as subordinates, are underrepresented in social media research. As Theocharis suggests [54], acts that are considered non-political can be more impactful than conventional forms of participation. In this section, we develop a theoretical basis to identify criteria that can contrast between a range of political personality types.

Early work done by Gamson [29] that highlighted the role discontent — lack of trust in the system—and efficacy — belief that individual actions matter—play in political behaviour provides a foundation for conceptualizing political personality types. Perceived efficacy is characterized by “judgments of how well one can execute courses of action required to deal with prospective situations” [55]. Trust, according to van der Meer [56], is a relational notion linking a subject (who trusts) with an object (that is trusted). Gamson [29] noted that the optimum combination for mobilization is high political efficacy and low political trust, and this combination indicates a belief that influence is possible and necessary. This notion, commonly known as the “Gamson Hypothesis”, has received considerable attention during the past several decades. Craig and Maggiotto [30] hypothesized that discontent is related to unconventional political behaviour among those who are dissatisfied with government policy and believe that they should be considered influential political actors. They conclude that the context has mobilized political discontent in the form of “benign” activities such as single-issue advocacy and tax revolt, and although these acts may not be revolutionary, they may have significant political consequences.

Similarly, Shingles [57] examined black consciousness and argued that policy-related participation is encouraged by political mistrust and internal efficacy. Shingles stresses the importance of understanding different types of political activity (e.g., allegiance to regime and country and influence on government and policy) and argues that efficacy can be conducive to high-initiative behaviour (i.e., activities that need active participation and high levels of effort), while mistrust is related to policy-related behaviour (i.e., efforts to influence government and policy). As Shingles notes, “strong sense of self-confidence and a deep suspicion regarding the willingness and ability of government to respond to their needs has proved to be an important catalyst, mobilizing blacks to seek to influence the policy process” [p.89].

Paige [23] examined the relationship between trust and efficacy, considering four types of political systems (democratic, traditional, totalitarian, and unstable). According to Paige, combinations of political trust and efficacy display differences among political actors (allegiants- high efficacy, high trust; subordinates- high trust, low efficacy; dissidents- high efficacy, low trust, and alienated- low trust, low efficacy) in these systems. While a responsive and non-coercive (democratic) regime requires little force due to allegiance, political alienation arises in a totalitarian regime that is unresponsive and coercive. This results in political allegiants being politically active but not radical as opposed to the alienated citizens that are resentful. In a traditional system characterised by a lack of trust but high efficacy, citizens can be seen as passive subordinates who are loyal to the existing political structure. Dissidence arises in an unstable (unresponsive and non-coercive) regime. Paige views this as critical, as revolutionary activity, including riot participation, can emerge. Paige claims that democratic, traditional, and totalitarian societies have been politically stable due to the nature of their constituency or the way they respond to interest groups.

Societies with dissidents, unlike the other three types, are highly unstable, as the mistrust and high efficacy can result in radical actions. The intricate connection between trust and efficacy provides a solid conceptual foundation to identify political personality types. Paige [23] argues that the behaviour of a given regime may move a system towards either democracy or totalitarianism. The above classification [29, p.23] allows considering every social media user as a political actor. For instance, lack of political engagement among a population of social media users may mean subordination or alienation. On the other hand, tendency to critique or support a political system as part of general social media use may indicate dissidence or allegiance. Accordingly, our effort to connect general social media use with political actor types is based on the premise that even the least politically active social media user plays an important role that helps understand the ways in which political activity is embedded in platforms. In particular, a comparative analysis among actors can reveal how general social media uses and gratifications may appeal to certain actor types.

Several researchers have used political efficacy and measures of political disaffection, such as political cynicism, to examine online political activity. For instance, Lee [58] argues that the use of online news and interaction with public agencies can increase internal political efficacy, while visiting websites of public agencies can positively affect internal efficacy. Chan and Guo [59] argue that the effect of Facebook use on political participation is moderated by political efficacy. An effort to directly apply Gamson’s proposition in the context of the Internet was made by Johnson, Kaye, and Kim [60], who addressed different web platforms, bulletin boards, chat rooms, portals, and blogs, and the effects of relying on these platforms. Their work shows that Internet users are equally divided into dissidents and assureds (allegiants). They claim that the results do not strongly support using the two categories (dissidents or assureds) as predictors of reliance on different platforms. However, a recent study conducted by Johnson and Kaye [61] shows that dissidents tend to use more polarizing media sources, such as radio talk shows and political blogs, while assureds avoid those sources. Johnson and Kaye claim that the use of online mainstream broadcast TV news predicts being an assured, while dissidence can be predicted by the use of alternative sources. These findings show that dissidents and allegiants can differ in terms of their use of online sources. However, arguably, any online platforms may have uses for both these parties.

Accordingly, the objective of this study is twofold: 1) identifying differences in social media uses and gratifications among potential dissidents, allegiants, subordinates, and the alienated, and 2) examining the extent to which political actor types can be discerned using social media uses and gratifications.

2 Method

Data from a sample of 313 United States citizens above the age of 18 (see Table 2 for sample statistics) was collected through Qualtrics survey service. The data collection team was advised to target the general population of the United States, rather than a selected group of strata from the population. They were also advised to continue data collection until a reasonable number of responses were received from different gender and ethnic groups, and individuals representing different age levels, education levels, political orientations, and party affiliations. Respondents were asked to reflect upon their use of Facebook and Twitter as they answered the survey questions. These two platforms were used as the main focus, as political engagement is common on Facebook and Twitter.

Table 2
Sample Composition

Demographic Variable Categories n %

Gender Male 131 41.9

Female 182 58.1

Age Between 18–25 years 23 7.3

Between 26–32 years 53 16.9

Between 33–40 years 64 20.4

Between 41–47 years 48 15.3

Above 47 years 125 39.9

Ethnicity Caucasian 236 75.4

African American 33 10.5

Asian 15 4.8

Native American 1 0.3

Pacific Islander 2 0.6

Hispanic 19 6.1

Other 7 2.2

Education College Freshman 17 5.4

Sophomore 20 6.4

Junior 12 3.8

Senior 26 8.3

Graduate Student 95 30.4

Other 143 45.7

Political Self-identification Liberal 80 25.6

Conservative 63 20.1

Conservative Democrat 32 10.2

Social Conservative 12 3.8

Moderate 68 21.7

Independent 58 18.5

Political Party Affiliation Democratic 155 49.5

Republican 85 27.2

Other 73 23.3

Social Media Use Several times day 193 61.7

About once a day 50 16

3–5 days a week 33 10.5

1–2 days a week 21 6.7

Every few weeks 12 3.8

Less often 4 1.3

Demographic Variable	Categories	n	%
Gender	Male	131	41.9
	Female	182	58.1
Age	Between 18–25 years	23	7.3
	Between 26–32 years	53	16.9
	Between 33–40 years	64	20.4
	Between 41–47 years	48	15.3
	Above 47 years	125	39.9
Ethnicity	Caucasian	236	75.4
	African American	33	10.5
	Asian	15	4.8
	Native American	1	0.3
	Pacific Islander	2	0.6
	Hispanic	19	6.1
	Other	7	2.2
Education	College Freshman	17	5.4
	Sophomore	20	6.4
	Junior	12	3.8
	Senior	26	8.3
	Graduate Student	95	30.4
	Other	143	45.7
Political Self-identification	Liberal	80	25.6
	Conservative	63	20.1
	Conservative Democrat	32	10.2
	Social Conservative	12	3.8
	Moderate	68	21.7
	Independent	58	18.5
Political Party Affiliation	Democratic	155	49.5
	Republican	85	27.2
	Other	73	23.3
Social Media Use	Several times day	193	61.7
	About once a day	50	16
	3–5 days a week	33	10.5
	1–2 days a week	21	6.7
	Every few weeks	12	3.8
	Less often	4	1.3

We used the revised MAIN model scale developed and validated by Rathnayake and Winter [39] to measure social media uses and gratifications. This scale includes 11 gratifications (Realism, Being There, Agency-enhancement, Community Building, Bandwagon, Filtering, Interaction, Activity, Responsiveness, Browsing, and Play) and was built on the Uses and Grats 2.0 framework [38, 52].

Definitions of each construct and gratifications are given in Table 1. The scale includes 66 questions re-written specifically to focus on social media. Cronbach’s Alpha values for all constructs were above 0.7 except for Filtering (Realism: 0.70, Coolness: 0.76, Being There: 0.80, Agency: 0.832, Community Building: 0.787, Filtering: 0.64, Activity: 0.77, Responsiveness: 0.835, Browsing: 0.88, Play: 0.8). A Confirmatory Factor Analysis (CFA) indicated reasonable fit, supporting the appropriateness of this scale (Model-fit: χ2 : 768.89, df: 389, p≤.001, GFI: 0.850, AGFI: 0.821, IFI: 0.924, TLI: 0.914, CFI: 0.923, RMSEA: 0.056).

Seven items (Table 3) developed by Johnson and Kaye [61] were used to measure trust and efficacy (Cronbach’s Alpha- political efficacy: 0.805; political trust: 0.654). A CFA that included political trust and efficacy showed adequate fit (Model-fit: χ2 : 42.35, df: 13, p:.000, IFI: 0.956, TLI: 0.928, CFI: 0.956, RMSEA: 0.085).

Table 3

Political Efficacy and Trust Measures

Construct	Items
	I consider myself well qualified to participate in politics
Efficacy	I feel I could do as good of a job in public office as most other people
	I think that I am better informed about politics and government than most people
	I feel that I have a pretty good understanding of the important political issues facing our country
	Most of our leaders are devoted to the service of our country
Trust in the government	Politicians never tell us what they really think
	I don’t think public officials care much about what people like me think

Subjects were categorised into political personality types based on the Gamson Hypothesis [29], Paige’s [23] conceptualisation of actor types, and related recent work [e.g., 60, 61] that used political trust and efficacy to identify dissidents, allegiants, subordinates, and alienated. While political allegiants indicate high trust and high efficacy, dissidents show low trust and high efficacy. Thus, political trust differentiates between allegiants and dissidents while these two groups share a sense of capability in political engagement. Those who trust the government, but indicate low efficacy, are identified as subordinates. Moreover, the alienated category shows low trust and low efficacy. Item totals were used to classify actors into personality categories. Following earlier work [e.g., 61], for constructs measured using four items (e.g., political efficacy), totals less than 12 were considered low and totals greater than 16 were categorized as high. Similarly, for constructs with three five-point Likert scale items (e.g., political trust), totals less than nine were categorised as low while totals higher than 12 were considered high. Accordingly, the sample included 11 dissidents, 37 allegiants, 50 subordinates, and 32 alienated subjects. A separate category was created for moderates to aid comparison of high and low categories with subjects that have moderate scores. The sample included 183 moderate subjects who did not fall into any of the main categories. This is a reasonable composition, as dissidents are political minorities who attempt to change the system, and it is typical that there are reasonably large groups that trust and support the existing system (allegiants). Moreover, it is possible that there are even larger groups that trust the system but are politically inactive (subordinates), and perhaps smaller inactive groups that do not trust the system (alienated). We used mean ranks and the Mann-Whitney U test to examine differences among political personality types due to unequal group sizes. We develop a multinomial logistic regression model to examine the relationship between predictors (uses and gratifications) and political personality types. Four extra variables (age, political party identification, political self-identification, frequency of social media use) were added to the regression model to improve classification accuracy. Political party identification was measured using a three-item scale (Democratic, Conservative, and Other) and five-item scale developed using the political typology suggested by the Pew Research Center [62] was used to measure political self-identification.

3 Results and discussion

Table 4 shows means and standard deviations for U&G constructs for each political personality type. While both dissidents and allegiants gravitated towards high mean values for some constructs (e.g., Realism, Coolness, Agency, Community Building, and Activity), they appeared to differ in terms of Filtering. However, both allegiants and dissidents had considerably different mean values from subordinates, moderates, and the alienated actors. Charts illustrated in Fig. 1 show corresponding mean values for each actor type.

Table 4
Means and Standard Deviations

Mean SD

Realism Moderates 2.44 0.839

Dissidents 3.39 0.998

Allegiants 3.26 1.250

Subordinates 2.27 0.838

Alienated 2.70 1.038

Total 2.57 0.973

Coolness Moderates 3.66 0.600

Dissidents 4.27 0.389

Allegiants 4.26 0.782

Subordinates 3.65 0.617

Alienated 3.51 0.568

Total 3.74 0.656

Being There Moderates 3.32 0.888

Dissidents 3.58 1.044

Allegiants 3.70 1.146

Subordinates 3.11 0.980

Alienated 3.27 0.759

Total 3.34 0.939

Agency Moderates 3.77 0.650

Dissidents 4.36 0.526

Allegiants 4.35 0.603

Subordinates 3.85 0.697

Alienated 3.69 0.829

Total 3.86 0.698

Community Building Moderates 3.75 0.698

Dissidents 4.09 0.634

Allegiants 4.12 0.759

Subordinates 3.57 0.774

Alienated 3.63 0.665

Total 3.77 0.726

Filtering Moderates 3.92 0.486

Dissidents 3.70 0.752

Allegiants 4.26 0.663

Subordinates 3.89 0.596

Alienated 3.64 0.551

Total 3.92 0.563

Activity Moderates 3.23 0.815

Dissidents 3.64 0.567

Allegiants 3.77 0.916

Subordinates 3.04 0.727

Alienated 2.99 0.760

Total 3.25 0.828

Responsiveness Moderates 3.21 0.746

Dissidents 3.24 1.184

Allegiants 3.85 0.915

Subordinates 2.96 0.755

Alienated 3.16 0.616

Total 3.24 0.807

Browsing Moderates 3.82 0.644

Dissidents 4.03 0.505

Allegiants 4.12 0.630

Subordinates 3.75 0.699

Alienated 3.62 0.615

Total 3.83 0.653

Play Moderates 3.47 0.856

Dissidents 3.79 0.688

Allegiants 3.94 0.984

Subordinates 3.49 0.817

Alienated 3.30 0.658

Total 3.52 0.855

		Mean	SD
Realism	Moderates	2.44	0.839
	Dissidents	3.39	0.998
	Allegiants	3.26	1.250
	Subordinates	2.27	0.838
	Alienated	2.70	1.038
	Total	2.57	0.973
Coolness	Moderates	3.66	0.600
	Dissidents	4.27	0.389
	Allegiants	4.26	0.782
	Subordinates	3.65	0.617
	Alienated	3.51	0.568
	Total	3.74	0.656
Being There	Moderates	3.32	0.888
	Dissidents	3.58	1.044
	Allegiants	3.70	1.146
	Subordinates	3.11	0.980
	Alienated	3.27	0.759
	Total	3.34	0.939
Agency	Moderates	3.77	0.650
	Dissidents	4.36	0.526
	Allegiants	4.35	0.603
	Subordinates	3.85	0.697
	Alienated	3.69	0.829
	Total	3.86	0.698
Community Building	Moderates	3.75	0.698
	Dissidents	4.09	0.634
	Allegiants	4.12	0.759
	Subordinates	3.57	0.774
	Alienated	3.63	0.665
	Total	3.77	0.726
Filtering	Moderates	3.92	0.486
	Dissidents	3.70	0.752
	Allegiants	4.26	0.663
	Subordinates	3.89	0.596
	Alienated	3.64	0.551
	Total	3.92	0.563
Activity	Moderates	3.23	0.815
	Dissidents	3.64	0.567
	Allegiants	3.77	0.916
	Subordinates	3.04	0.727
	Alienated	2.99	0.760
	Total	3.25	0.828
Responsiveness	Moderates	3.21	0.746
	Dissidents	3.24	1.184
	Allegiants	3.85	0.915
	Subordinates	2.96	0.755
	Alienated	3.16	0.616
	Total	3.24	0.807
Browsing	Moderates	3.82	0.644
	Dissidents	4.03	0.505
	Allegiants	4.12	0.630
	Subordinates	3.75	0.699
	Alienated	3.62	0.615
	Total	3.83	0.653
Play	Moderates	3.47	0.856
	Dissidents	3.79	0.688
	Allegiants	3.94	0.984
	Subordinates	3.49	0.817
	Alienated	3.30	0.658
	Total	3.52	0.855

Fig. 1

Political Actor Category and Uses and Gratifications (Mean Values).

The results of the Mann-Whitney U test (Table 5) show that dissidents and allegiants did not differ from each other in terms of their perception of uses and gratifications, except for Filtering. Allegiants had a significantly higher mean rank than dissidents (mean ranks- dissidents: 136.32, allegiants: 213.55, Mann-Whitney U: 111.00, p < .05) for their perception of Filtering. However, both these groups differed significantly from other groups. For instance, perception of Realism, Coolness, Agency, Community Building, and Activity was significantly different between dissidents and subordinates. Similarly, dissidents and the alienated were different from each other in terms of their perception of Coolness, Agency, Community Building, and Activity. Although dissidents did not differ from subordinates and allegiants for some constructs (e.g., Being There, Filtering, Responsiveness, and Browsing), differences between political allegiants and the two groups with low political efficacy (subordinates and the alienated) were more prominent. For instance, political allegiants and the alienated were significantly different from each other for all uses and gratifications constructs. Similarly, allegiants and subordinates were different from each other for all constructs except Being There. Accordingly, we have convincing evidence to argue that, from the perspective of uses and gratifications, dissidents and allegiants do not differ much from each other, but do differ from other groups, such as subordinates and the alienated. This similarity may relate to high efficacy among dissidents and allegiants. In other words, these results show that social media constitute a ‘level playing field’ for those who believe in the significance of individual actions in politics. Mean ranks given in Table 4 and charts that provide a comparative perspective of mean values (Fig. 1) also showed that potential dissidents and allegiants perceive uses and gratifications differently than the other two groups. This shows that social media can gratify dissidents and allegiants more than subordinates and the alienated.

Table 5

Mann-Whitney U Test Results for Group Differences

	Test 1:		Test 2:		Test 3:		Test 4:		Test 5:
	Dissidents vs. Allegiants		Dissidents vs. Subordinates		Allegiants vs. Subordinates		Dissidents vs. Alienated		Allegiants vs. Alienated
	Mann-Whitney U	p	Mann-Whitney U	p	Mann Whitney U	p	Mann-Whitney U	p	Mann-Whitney U	p
Realism	200.500	0.941	106.500	0.001	488.000	0.000	111.000	0.068	413.500	0.031
Coolness	177.000	0.506	107.000	0.001	401.500	0.000	48.000	0.000	200.000	0.000
Being There	182.500	0.604	206.500	0.195	607.000	0.006	140.500	0.317	401.000	0.021
Agency	200.500	0.940	155.000	0.022	533.000	0.001	77.000	0.004	287.000	0.000
Community	187.000	0.677	161.500	0.031	464.500	0.000	102.500	0.036	288.500	0.000
Building
Filtering	111.000	0.021	241.500	0.515	574.500	0.002	151.500	0.472	249.500	0.000
Activity	166.500	0.359	139.500	0.010	464.500	0.000	90.000	0.015	289.000	0.000
Responsiveness	138.000	0.105	226.500	0.359	388.000	0.000	162.000	0.693	286.500	0.000
Browsing	176.500	0.500	222.500	0.316	648.000	0.016	112.000	0.068	333.000	0.002
Play	163.000	0.312	208.500	0.206	592.500	0.004	107.500	0.054	303.000	0.000

^aGrouping Variable: Actor Category.

A multinomial logistic regression model was used to examine the relationship between social media uses and gratifications and political personality types. Age of the respondents, frequency of social media use, political party affiliation, and political self- identification were included to improve the model. The model with the predictors indicated reasonable fit (χ²: 702.874, df: 424, p:0.00, Cox and Snell R²:0.894, Nagelkerke R²:0.981, McFadden R²:0.928). Likelihood ratio test results (Table 6) showed that seven out of thirteen variables (i.e., age, political self-identification, frequency of social media use, Coolness, Filtering, Activity, and Play) included in the final model were significant (p < .05). Classification results (Table 7) show that the predictors can classify users into political personality types with more than 98 percent accuracy. A second multinomial logistics model that did not include political party affiliation and self-identification classified the sample into political personality categories with 92.3% percent accuracy (moderates: 94.5%, dissidents: 100%, allegiants: 100%, subordinates: 74%, alienated: 96.9%) (Model Fit- Log Likelihood: 107.191, χ²: 646.13, df: 408, p:0.00). This demonstrates that a model that does not include predictors related to political attributes of respondents can still achieve high classification accuracy.

Table 6

Multinomial Logistic Regression- Likelihood Ratio Tests

	Model Fitting Criteria –2 Log Likelihood of Reduced Model	Likelihood Ratio Tests
Effect		Chi-Square	df	Sig.
Intercept	54.607	0.000	0
Age	97.094	42.487	4	0.000
Political party identification	58.304	3.697	4	0.449
Political self-identification	67.527	12.921	4	0.012
Frequency of social media use	67.213	12.606	4	0.013
Realism	95.978	41.371	48	0.739
Coolness	359.907	305.301	40	0.000
Being There	86.331	31.724	48	0.966
Agency	82.797	28.190	44	0.969
Community Building	94.569	39.962	44	0.645
Filtering	236.856	182.249	40	0.000
Activity	186.211	131.604	48	0.000
Browsing	101.249	46.643	48	0.529
Play	124.599	69.993	48	0.021

Table 7

Classification Results

	Predicted
Observed	Moderates	Dissidents	Allegiants	Subordinates	Alienated	Percent Correct
Moderates	181	0	0	1	1	98.9%
Dissidents	0	11	0	0	0	100.0%
Allegiants	0	0	37	0	0	100.0%
Subordinates	3	0	0	47	0	94.0%
Alienated	0	0	0	0	32	100.0%
%	58.8%	3.5%	11.8%	15.3%	10.5%	98.4%

The above results allow us to make two claims: 1) potential allegiants and dissidents are driven by a similar set of social media uses and gratifications as opposed to political subordinates and the alienated, and 2) social media uses and gratifications can be used to classify political personality types with considerably high accuracy. This answers the general question, ‘can data related to how and why users are driven to use social media accurately reveal political differences?’, in the affirmative. Our results also show that social media can provide more gratifications for potential dissidents and allegiants, ‘favouring’ personality types with high political efficacy. We will discuss below the implications of these findings. Starting from early work that discussed the impact of the Internet on political participation in general [e.g., 63] previous studies have reached into numerous dimensions of online politics. For instance, studies that examine dissidence in different contexts, such as political uprisings and oppression [e.g., 27, 53, 64], provide ample evidence to understand the role social media play in enabling political activity. These studies approach online political dissidence from an issue-based perspective, focusing on specific contexts. In contrast, our aim was to provide a more general view of how political behaviour is connected to social media use. Our approach allows examination of less-oppressive and non-issue-specific political contexts and takes different political actor types, including passive actors, into account.

As Gillespie [65] notes, choices made by platforms regarding the nature of content that can appear, organization and monetization of such content, and prohibition of some information are crucial interventions into the public discourse. However, platforms may not favour certain political ideologies or orientations. Given the vast amount of studies that examine alternative politics such as protest action [e.g., 7, 8, 53, 66–69], social media platforms may seem favourable venues for alternative politics, particularly dissidence. However, arguably, the types of affordances identified by social media scholars (e.g., persistence, replicability, scalability, and searchability, as discussed by boyd [70]), do not appeal in particular to a specific political group. Similarly, informational, expressive, and relational uses of social media that are positively linked to engagement [1] can appeal to users with different political orientations. Moreover, social media uses and gratifications identified by previous studies (e.g., information seeking, entertainment, social interaction, self-status seeking) cannot be limited to specific political ideologies. Therefore, although accessibility of platforms may encourage dissident politics that is less likely to be promoted in broadcast media, the potential for allegiant or pro-regime politics on platforms is not negligible. For instance, as Pearce [16] argues, despite the optimistic view towards the democratic potential of the Internet, social media can make control easy for authoritarian regimes. Similarities in uses and gratifications between potential dissidents and allegiants that we identified above may, therefore, reflect the potential of platforms to enable both pro and anti-system politics.

These results can also explain the intricate connection between political efficacy and online political engagement. Mean values for political efficacy (based on a five-point Likert Scale) were similar between potential dissidents and allegiants (mean = dissidents: 4.36, allegiants: 4.41) and both groups had a mean rank of 24.5 (the Mann-Whitney U test was not significant). As both potential dissidents and allegiants have high political efficacy, they may perceive some social media uses and gratifications more than those who have less political efficacy. This is possible, as social media is characterised by user-generated content and interaction. Differences in uses and gratifications between personality types with high and low political efficacy was clearly noticeable in the mean ranks given in Table 4, as dissidents and allegiants had higher mean ranks for all uses and gratifications than subordinates and the alienated. Partial correlations between Political Trust, Efficacy, and uses and gratifications (Table 8) indicted that, while only one gratification (Filtering) is significantly related to Political Trust, Political Efficacy had significant correlations with all the uses and gratifications variables. This confirms that social media tend to gratify users with high political efficacy more than others. In other words, our results show that social media provide higher levels of gratifications for those who believe in the significance of individual political activity. This observation is consistent with the study conducted by Velasquez and LaRose [71] who highlighted the correspondence between perceived efficacy and levels of agency at which political engagement occurs. Political efficacy influences many aspects of human functioning, such as strategic thinking, optimism or pessimism, choice of course of action, goals, effort put to certain endeavours, perseverance, and resilience [72]. Bandura notes that shared beliefs in people’s collective power to achieve desired outcomes can be seen as a key element of collective agency. Arguably, the ability of social media platforms to allow collective activity may correlate with collective efficacy of social media users. However, effects of collective efficacy on social media uses and gratifications should be examined further as it is beyond the scope of the current study.

Table 8

Partial Correlations among Political Trust, Efficacy, and Uses and Gratifications

	Trust	Efficacy
Realism	–0.014	0.214^**
Coolness	0.073	0.250^**
Being There	–0.037	0.154^**
Agency	0.066	0.152^**
Community Building	0.064	0.226^**
Filtering	0.180^**	0.154^**
Activity	0.055	0.259^**
Responsiveness	0.045	0.190^**
Browsing	0.096	0.172^**
Play	0.101	0.167^**

^**Correlation is significant at 0.01 level, ^*Correlation is significant at 0.05 level, Control variable: frequency of social media use.

The above results can offer valuable insight for both governments and platforms, especially in terms of developing policies to encourage democratic engagement. As mentioned previously, social media use is not limited to political motives. Therefore, in contexts where the population is diverse in terms of political position, different groups, including those who have dissident or allegiant tendencies, may find social media sites gratifying. However, as both dissidents and allegiants share a similar set of gratifications, measures taken to encourage or restrict engagement of one group may affect the other similarly. Paige [23] explains that repressive responses to control dissidence may transform an unstable regime into a totalitarian regime. Use of social media for such control may exacerbate the issue as such measures may negatively affect political allegiants. For instance, measures such as access control and limiting opportunity for interaction adversely affect both dissidents and allegiants as they reduce gratifications such as community building, agency, and activity.

This may result in dissatisfaction among allegiants, minimising support for government. Paige describes democratic, traditional, and totalitarian systems as stable, as opposed to unstable regimes characterised by dissidence. In stable, yet democratically unhealthy systems, such as totalitarian political contexts, social media can be used to encourage political engagement among both potential dissidents and allegiants.

4 Conclusion

The findings discussed above reveal a nuanced relationship between political personality types and social media uses and gratifications. Social media are double-edged swords that not only afford, but also gratify, individuals engaged in anti- as well as pro-system politics. However, they are not necessarily ‘level playing fields’, as they tend to provide more gratifications for those who believe in the value of individual political activity. While this study sheds light on the topic, further work is necessary to examine factors that mediate and moderate the relationships between social media uses and gratifications and political personality types. In particular, work that examines political motivations can help shed light on the topic. Unequal group sizes within the sample also demands similar work in different contexts.

Our classification above reveals the possibility of discerning political differences using data related to social media consumption. This could prove to be problematic, as it helps uncover individuals that support or oppose existing political systems, leading to potential harms. However, this work should not be understood as an attempt to expose political actors, those who have mobilization potential in particular. Instead, our work should be understood as empirical evidence as to the possibility of such exposure using survey data. Nefarious use of such potential indicates democratic decay, and this demands strong policy frameworks for platform use, analytics, and micro targeting.

References

Skoric

, Zhu

, Goh

, Pang

. Social media and citizen engagement: A meta-analytic review. New Media & Society. 2016;18(9):1817–39. https://doi.org/10.1177/1461444815616221

Bailo

, Vromen

. Hybrid social and news media protest events: from #MarchinMarch to #BusttheBudget in Australia. Information Communication and Society. 2017;20(11):1660–79. https://doi.org/10.1080/1369118X.2016.1252410

Harlow

, Benbrook

. How #Blacklivesmatter: exploring the role of hip-hop celebrities in constructing racial identity on Black Twitter. Information Communication and Society. 2017;4462:1–17. https://doi.org/10.1080/1369118X.2017.1386705

Jackson

S J

, Foucault Welles

. #Ferguson is everywhere: initiators in emerging counterpublic networks. Information Communication and Society. 2016;19(3):397–418. https://doi.org/10.1080/1369118X.2015.1106571

Juris

J S

. Reflections on #Occupy Everywhere: Social media, public space, and emerging logics of aggregation. American Ethnologist. 2012;39(2):259–79. https://doi.org/10.1111/j.1548-1425.2012.01362.x

van Laer

, van Aelst

. Internet and social movement action repertoires. Information, Communication & Society. 2010;13(8):1146–71. https://doi.org/10.1080/13691181003628307

Lim

. Clicks, cabs, and coffee houses: Social media and oppositional movements in Egypt, 2004-2011. Journal of Communication. 2012;62(2):231–48. https://doi.org/10.1111/j.1460-2466.2012.01628.x

Meraz

, Papacharissi

. Networked gatekeeping and networked framing on #Egypt. International Journal of Press/Politics. 2013;18(2):138–66. https://doi.org/10.1177/1940161212474472

Parsloe

, Holton

. #Boycottautismspeaks: communicating a counternarrative through cyberactivism and connective action. Information Communication and Society. 2018;21(8):1116–33. https://doi.org/10.1080/1369118X.2017.1301514

10.

Poell

. Social media and the transformation of activist communication: exploring the social media ecology of the 2010 Toronto G20 protests. Information, Communication & Society. 2013;17(6):716–31. https://doi.org/10.1080/1369118X.2013.812674

11.

Tremayne

. Anatomy of Protest in the Digital Era: A Network Analysis of Twitter and Occupy Wall Street. Social Movement Studies. 2013:Taylor & Francis. https://doi.org/10.1080/14742837.2013.830969

12.

Vicari

, Iannelli

, Zurovac

. Political hashtag publics and counter-visuality: a case study of #fertilityday in Italy. Information, Communication & Society. 2018:1–20. https://doi.org/10.1080/1369118X.2018.1555271

13.

Youmans

, York

. Social Media and the activist toolkit: User agreements, corporate interests, and the information infrastructure of modern social movements. Journal of Communication. 2012;62(2):315–29. https://doi.org/10.1111/j.1460-2466.2012.01636.x

14.

Barberá

, Jost

, Nagler

, Tucker

, Bonneau

. Tweeting from left to right: Is online political communication more than an echo chamber? Psychological Science. 2015:1–12. https://doi.org/10.1177/0956797615594620

15.

Engesser

, Ernst

, Esser

, Büchel

. Populism and social media: How politicians spread a fragmented ideology. Information Communication and Society. 2017;20(8):1109–26. https://doi.org/10.1080/1369118X.2016.1207697

16.

Pearce

. Democratizing kompromat: The affordances of social media for state-sponsored harassment. Information Communication and Society. 2015;18(10):1158–74. https://doi.org/10.1080/1369118X.2015.1021705

17.

Groshek

, Koc-Michalska

. Helping populism win? Social media use, filter bubbles, and support for populist presidential candidates in the 2016 US election campaign. Information Communication and Society. . 2017;20(9):1389–407. https://doi.org/10.1080/1369118X.2017.1329334

18.

Russo

, Amnå

. The personality divide: Do personality traits differentially predict online political engagement?Social Science Computer Review. 2016;34(3):259–77. https://doi.org/10.1177/0894439315582487

19.

Bimber

, Cunill

, Copeland

, Gibson

. Digital media and political participation: The moderating role of political interest across acts and over time. Social Science Computer Review. 2015;33(1):21–42. https://doi.org/10.1177/0894439314526559

20.

Boulianne

. Online news, civic awareness, and engagement in civic and political life. New Media and Society. 2016;18(9):1840–56. https://doi.org/10.1177/1461444815616222

21.

Malik

, Dhir

, Nieminen

. Uses and gratifications of digital photo sharing on Facebook. Telematics and Informatics. 2016;33(1):129–38. https://doi.org/10.1016/j.tele.2015.06.009

22.

Park

, Kee

, Valenzuela

. Being immersed in social networking environment: Facebook groups, uses and gratifications, and social outcomes. Cyberpsychology & Behavior. 2009;12(6):729–33.

23.

Paige

. Political orientation and riot participation. American Sociological Review. 1971;36(5):810–20.

24.

Wolfsfeld

. INFOCORE definitions: “Political Actors.” 2015. Retrieved from https://www.infocore.eu/results/definitions/

25.

Al-Ani

, Mark

, Semaan

. Blogging in a region of conflict: Supporting transition to recovery. Proceedings of the SIGCHI Conference on Human Factors in Computing Systems, 2010:1069–78. Retrieved from https://dl.acm.org/citation.cfm?id=1753485

26.

Mark

, Semaan

. Expanding a country’s borders during war: The Internet war diary. In Proceedings of the 2009 international workshop on Intercultural collaboration. 2009:3–12.

27.

Marzouki

, Skandrani-Marzouki

, Béjaoui

, Hammoudi

, Bellaj

. The contribution of Facebook to the 2011 Tunisian revolution: a cyberpsychological insight. Cyberpsychology, Behavior and Social Networking. 2012;15(5):237–44. https://doi.org/10.1089/cyber.2011.0177

28.

Rathnayake

, Winter

. Examining the link between social media uses and gratifications, and political tolerance and dogmatism. Policy and Internet. 2017;9(4):444–66. https://doi.org/10.1002/poi3.157

29.

Gamson

. Power and discontent. 1968: The Dorsey Press.

30.

Craig

, Maggiotto

. Political discontent and political action. The Journal of Politics. 1981;43(2):514–22.

31.

Kaye

, Johnson

. Online and in the know: Uses and gratifications of the web for political information. Journal of Broadcasting & Electronic Media. 2002;46(1):54–71. https://doi.org/10.1207/s15506878jobem4601

32.

Vaidhyanathan

. Anti-social media: How Facebook disconnects us and undermines democracy. 2018: Oxford University Press.

33.

Ruggiero

. Uses and Gratifications Theory in the 21st century. Mass Communication and Society. 2000; 3(1):3–37. https://doi.org/10.1207/S15327825MCS0301_02 [18] Russo S, Amnå E. The personality divide: Do personality traits differentially predict online political engagement? Social Science Computer Review. 2016; 34(3):259–77. https://doi.org/10.1177/0894439315582487

34.

Alhabash

, Chiang

, Huang

. MAM & U&G in Taiwan: Differences in the uses and gratifications of Facebook as a function of motivational reactivity. Computers in Human Behavior. 2014;35:423–30. https://doi.org/10.1016/j.chb.2014.03.033

35.

Chen

. Tweet this: A uses and gratifications perspective on how active Twitter use gratifies a need to connect with others. Computers in Human Behavior. 2011;27(2):755–62. https://doi.org/10.1016/j.chb.2010.10.023

36.

Chen

H-T

, Kim

. Problematic use of social network sites: The Interactive relationship between gratifications sought and privacy concerns. Cyberpsychology, Behavior and Social Networking. 2013;16(11):806–13. https://doi.org/10.1089/cyber.2011.0608

37.

Smock

, Ellison

, Lampe

, Wohn

. Facebook as a toolkit: A uses and gratification approach to unbundling feature use. Computers in Human Behavior. 2011;27(6):2322–9. https://doi.org/10.1016/j.chb.2011.07.011

38.

Sundar

, Limperos

. Uses and grats 2. New gratifications for new media. Journal of Broadcasting & Electronic Media. 2013;57(4):504–25. https://doi.org/10.1080/08838151.2013.845827

39.

Rathnayake

, Winter

. Carrying forward the uses and grats 2. 0 agenda: An affordance-driven measure of social media uses and gratifications. Journal of Broadcasting & Electronic Media. 2018;62(3):371–89. https://doi.org/10.1080/08838151.2018.1451861

40.

Katz

, Blumler

, Gurevitch

. Uses and gratifications research. The Public Opinion Quarterly. . 1973;37(4):509–23.

41.

Haridakis

. Uses and gratifications: A social and psychological perspective of media use and effects. In The International Encyclopedia of Media Studies. 2013: Blackwell Publishing Ltd., pp. 1–22).

42.

Papacharissi

. Uses and gratifications. In D. W. Stacks & M. B. Salwen (Eds.), An integrated approach to communication theory and research. 2009: Routledge, 137–52.

43.

Anderson

. The uses and gratifications of online care pages: A study of CaringBridge. Health Communication. 2011;26(6):546–59. https://doi.org/10.1080/10410236.2011.558335

44.

Cortese

, Rubin

. Uses and gratifications of television home shopping. Atlantic Journal of Communication. 2010;18:89–109. https://doi.org/10.1080/15456870903554924

45.

Gudelunas

. There’s an app for that: The uses and gratifications of online social networks for gay men. Sexuality & Culture. 2012;16(4):347–65. https://doi.org/10.1007/s12119-012-9127-4

46.

Kim

. Scan and click: The uses and gratifications of social recommendation systems. Computers in Human Behavior. 2014;33:184–91. https://doi.org/10.1016/j.chb.2014.01.028

47.

Shao

. Understanding the appeal of user-generated media: A uses and gratifications perspective. Internet Research. 2009;19(1):7–25. https://doi.org/10.1108/10662240910927795

48.

J-H

, Wang

S-C

, Tsai

H-H

. Falling in love with online games: The uses and gratifications perspective. Computers in Human Behavior. 2010;26(6):1862–71. https://doi.org/10.1016/j.chb.2010.07.033

49.

Ancu

, Cozma

. MySpace politics: Uses and gratifications of befriending candidates. Journal of Broadcasting & Electronic Media. 2009;53(December):567–83. https://doi.org/10.1080/08838150903333064

50.

Evans

, Pearce

, Vitak

, Treem

. Explicating affordances: A conceptual framework for understanding affordances in communication research. Journal of Computer-Mediated Communication. 2017;22(1):35–52. https://doi.org/10.1111/jcc4.12180

51.

Gibson

. The theory of affordances. The ecological approach to visual perception. 1986: Lawrence Erlbaum Associates.

52.

Sundar

. The MAIN model: A heuristic approach to understanding technology effects on credibility. In M. J. Metzger & A. J. Flanagin (Eds.), Digital media, youth, and credibility. 2008: The MIT Press, 73–100. https://doi.org/10.1162/dmal.9780262562324.073

53.

Al-Ani

, Mark

, Chung

, Jones

. The Egyptian blogosphere: A counter- narrative of the revolution. In Proceedings of the ACM 2012 conference on Computer Supported Cooperative Work, 2012:17–26.

54.

Theocharis

. The Conceptualization of Digitally Networked Participation. 2015;(October). https://doi.org/10.1177/2056305115610140

55.

Bandura

. Self-efficacy mechanism in human agency. American Psychologist. 1982;37(2):122–47. https://doi.org/10.1016/0006-8993(86)91535-0

56.

van der Meer

TWG

. Political trust and the “crisis of democracy. ” In Oxford Research Encyclopedia of Politics. 2017;1:1–22. https://doi.org/10.1093/acrefore/9780190228637.013.77

57.

Shingles

. Black consciousness and political participation: The missing link. The American Political Science Review. 1981;75(1):76–91. https://doi.org/10.2307/1962160

58.

Lee

. Effects of internet use on college students’ political efficacy. Cyberpsychology & Behavior: The Impact of the Internet, Multimedia and Virtual Reality on Behavior and Society. 2006;9(4):415–422. https://doi.org/10.1089/cpb.2006.9.415

59.

Chan

, Guo

. The role of political efficacy on the relationship between Facebook use and participatory behaviors: A comparative study of young American and Chinese adults. Cyberpsychology, Behavior and Social Networking. 2013;16(6):460–463. https://doi.org/10.1089/cyber.2012.0468

60.

Johnson

T J

, Kaye

B K

, Kim

. Creating a web of trust and change: Testing the Gamson Hypothesis on politically interested internet users. Atlantic Journal of Communication. 2010;18:259–279. https://doi.org/10.1080/15456870.2010.521474

61.

Johnson

T J

, Kaye

B K

. Putting out fire with gasoline: Testing the Gamson Hypothesis on media reliance and political activity. Journal of Broadcasting & Electronic Media. 2013;57(4):456–481. https://doi.org/10.1080/08838151.2013.845825

62.

Pew Research Center. Beyond red vs. blue: Republicans divided about role of government - democrats by social and personal values. 2005. Retrieved October 22, 2018, from Beyond Red vs. Blue%0ARepublicans Divided About Role of Government - Democrats by Social and Personal Values

63.

Bimber

. The Internet and political transformation: Populism, community, and accelerated pluralism. Polity. . 1998;31(1):133–160.

64.

Shklovski

, Valtysson

. Secretly political: Civic engagement in online publics in Kazakhstan. Journal of Broadcasting & Electronic Media. 2012;56(3):417–433.

65.

Gillespie

. The politics of “platforms. ” New Media & Society. 2010;12:347–364. https://doi.org/10.1177/1461444809342738

66.

Liu

. Tweeting, re-tweeting, and commenting: Microblogging and social movements in China. Asian Journal of Communication. 2015;25(6):567–583. https://doi.org/10.1080/01292986.2015.1013971

67.

Peña-López

, Congosto

, Aragón

. Spanish Indignados and the evolution of the 15M movement on Twitter: Towards networked para-institutions. Journal of Spanish Cultural Studies. 2014;(August):1–28. https://doi.org/10.1080/14636204.2014.931678

68.

Theocharis

, Lowe

, van Deth

J W

, García-Albacete

. Using Twitter to mobilize protest action: Online mobilization patterns and action repertoires in the Occupy Wall Street, Indignados, and Aganaktismenoi movements. Information, Communication & Society. 2014;18(2):202–220. https://doi.org/10.1080/1369118X.2014.948035

69.

Wojcieszak

. Will politics be tweeted? New media use by Iranian youth in 2011. New Media & Society. 2014;16(1):91–109. https://doi.org/10.1177/1461444813479594

70.

boyd d. Social Network Sites as networked publics: Affordances, dynamics, and implications. In Z. Papacharissi (Ed.), A networked self: Identity, community, and culture on social network sites. 2011: New York and London: Routledge Taylor & Francis Group, 39–58. https://doi.org/10.4324/9780203876527

71.

Velasquez

, LaRose

. Youth collective activism through social media: The role of collective efficacy. New Media and Society. 2015;17(6):899–918. https://doi.org/10.1177/1461444813518391

72.

Bandura

. Exercise of human agency through collective efficacy. Current Directions in Psychological Research. 2000;9(3):75–8.