--- title: " " output: bookdown::pdf_document2: default header-includes: - \usepackage{float} - \usepackage{sectsty} --- # Introduction      The rapid evolution of media technology has profoundly reconstituted our interactional experiences and transformed the global political landscape. In recent years, social media platforms have emerged as essential intermediaries for politicians to exert political and societal influence. Not only has media technology facilitates sharing of information, thoughts, and ideas, but it has also fermented political insularity, incivility, and polarization. \par      One of the major political consequences of technological advances is the spread of disinformation. Since the 2016 U.S. presidential election, American democracy has been constantly buffeted under the weight of political disinformation. The prevalence and effects of bogus information on social media have raised concerns about how political falsehoods deteriorate ideological segregation and political polarization in the American society. Studies of proliferation of multifarious forms of false information, such as fake news, misconceptions, conspiracy theories, rumors, have been gaining ground in political science and communication research. Despite the extensive literature on the effects of disinformation on political attitudes, it seems that relatively insufficient attention has been devoted to the mechanisms of behavioral patterns exhibited by individuals encountering political disinformation.\par      This study is an attempt to examine individual disinformation sharing behavior on social media. The overarching research questions of this thesis addresses whether individuals share disinformation in alignment with their party affiliation, and if their levels of political knowledge, social media engagement, and other factors predict their sharing behaviors. I start from the review of pertinent literature emanating primarily from political science, social psychology, communication, and information research.\par      In reviewing the literature, I seek to relate my research to theoretical frameworks and arguments tackling the following research problems: How disinformation is conceptualized and operationalized; What motivates individuals to believe disinformation. \par ## Conceptualization and operationalization of disinformation      Disinformation has long been a leitmotif in human history. However, only since recent years has technology enabled the exponential development of dataflow and fueled the spread of bogus news and computational propaganda. In the United States, social media has been tooled to distribute fear-mongering conspiracies as well as to deploy disinformation campaigns blended with truth, lies and sincere beliefs.      The advent of information technology has revolutionized the way individuals produce, receive and consume information. According to a Pew report [@shearer2017], social networking sites, such as Facebook, Twitter, and YouTube, have become the main sources of news and information consumption for American adults. With the explosion of information and data, recent research has begun to explore the patterns and consequences of fake news and disinformation sharing in social media.      To gauge the influences of disinformation on democracy and elections, I first present how disinformation is conceptualized in the literature, which is defined by information and political scientists as falsified and distorted information intended to deceive or mislead [@stahl2006difference; @karlova2013social;@guess2020misinformation;@karlova2011notes;@losee1997discipline; @zhou2007ontology]. Accordingly, disinformation, fake news, conspiracy theories, rumors, propaganda, among other types of bogus information, are subsumed under this definition. Building upon this line of foundational work, I define disinformation in my thesis as a form of information fabricated with a deceptive or misleading intent.      Political scientists have been exploring the influences of disinformation in the conduct of politics from various theoretical perspectives, with a particular focus on public opinion research. The ubiquity of social media platforms amplifies misconceptions, false claims and fabricated information, which has been threatening the very foundations of democracy. @kuklinski2000misinformation argue that misinformation is a barrier to an informed citizenry as it has the potential “to skew aggregate opinion ( p. 22).” Disinformation consumption promotes false beliefs, political polarization, and deterioration of public trust of the democratic system [@ognyanova2020misinformation]. In addition, prior studies find that disinformation decreases trust in media organizations [@hulcoop2017tainted]. Individuals who visit websites or social media platforms fraught with falsified or unverified content are more likely to have biased political judgments and perceptions of facts [@berinsky2017rumors;@guess2020misinformation]. To recapitulate, not only disinformation undermines an informed citizenry, but it also erodes our confidence in electoral democracy.      In addition, a burgeoning literature is devoted to discussing disinformation as a cyber security threat. Foreign intelligent agencies and malicious actors equipped with advanced technology may deploy disinformation campaigns to disseminate political rumors about ballot fraud and election manipulation, causing voters to question the integrity of democratic voting systems and even the election’s legitimacy. In fact, political scientists have observed and documented waves of cyber-attacks targeting democracies [@lim2019burned;@stukal2017detecting]. On another note, the affordances of technology and disinformation campaigns have also become the tool for authoritarian states to suppress political dissidents and further buttress political support for authoritarian regime [@huang2017war;@livingston2019future].      Another line of research pertaining to (dis)information acceptance is Selective Exposure Theory. Defined as “any systematic bias in audience composition [@freedman1965selective],” Selective Exposure Theory describe the phenomenon in which individuals are “(self)exposed to disproportionately more supportive than nonsupportive information [@freedman1965selective].” Over the past few decades, as the array of media choices available has exploded, social scientists have started to focus their attention on how selective media exposure shapes one’s political attitude and behavior. It is noted that individuals who are exposed to an information environment replete with biased content tend to opt for media choices that conform to or reinforce their pre-existing political views while purposefully avoiding media content incongruent with their beliefs [@arceneaux2012cognitive;@garrett2009echo;@iyengar2008selective;@knobloch2009looking;@messing2014selective;@jomini2007media;@stroud2008media;@stroud2010polarization;@valentino2009efficacy;@weeks2015emotions]. Though partisan media and information sources do not necessarily result in political polarization [@prior2013media], these studies suggest that selective exposure to particular types of information may shape one’s political attitude and behavior. However, the existing literature appears to dominantly focus on the effects of disinformation, but pays relatively scant attention to behavioral patterns of disinformation sharing in a social media environment. Therefore, in this thesis, I address this research gap with a survey experiment designed to explore partisan disinformation sharing behavior on social media platforms.      Theories in social psychology also shed lights on understanding the mechanisms of disinformation dissemination. Social psychologists examine political disinformation at a relatively micro level as they tend to focus on disinformation’s effects on individual cognitive infrastructure. Recent research shows that emotional reactions, such as fear, anger, disgust, and sadness have the potential to trigger individuals’ favorable responses to bogus messages and claims as sentimental attachment and affections may deprive their cognitive capabilities to discern authentic information from dubious one [@porter2010prospective;@rosenzweig2020misinformation;@van2014power;@weeks2015emotions]. Other psychological factors that may drive individuals to fall prey to fake information are, but not limited to, perceptions of accuracy [@pennycook2019lazy;@pennycook2020implied], quality of information sources [@dias2020emphasizing;@pennycook2019fighting], prior exposure [@pennycook2018prior], and political beliefs [@swire2017processing;@pennycook2019lazy;@thorson2016belief].      Additionally, economists have provided a fresh theoretical perspective in which they treat the media system as a “supply and demand” market. On one hand, proliferation of disinformation can be seen as a result of malfunction of the information supply chain, where fake news and deceitful information have been constantly produced by malicious actors, cyber criminals, and biased media organizations to pollute the information ecosystem. On the other hand, a lack of media literacy and the increasing political divide drive the demand side of the disinformation market, in which individuals seek partisan and biased information to endorse their political beliefs [@allcott2019trends]. @allcott2017social put forward a model illustrating the market logic of disinformation dissemination from both the supply and demand sides. In terms of provisions of disinformation, content generators provide disinformation with a profit motive to drive more traffic to their websites; ergo they may deliberately fabricate ideologically biased disinformation to accommodate partisans. On the demand side, individuals desire such fabricated information and willingly expose themselves to it and consider it as a heuristic cue to strengthen their group and partisan identities. Simply put, individuals may be motivated to share disinformation for different reasons. It is, therefore, vital to investigate what factors shape disinformation sharing behavior. ## Partisanship and party identification      Partisan identity dictates political behaviors and attitudes [@gerber2010party]. Previous research suggests that party affiliation or partisanship plays a crucial role in one’s political assessments. Citizens use partisan cues to evaluate the accuracy and reliability of the information sources from political figures, organizations, and media, and to process political information in order to understand the social and political structure of society. The partisanship literature traces its roots to the Michigan political scientists @belknap1951political's work on the sociological mechanism of political attitudes. Following in their footsteps, @campbell1980american put forward the “Michigan model” in one of the classic theoretical studies of political behavior - *The American Voter* to illustrate how the public forms political preferences. As one of the key clusters of an individual’s political inclinations, a person’s party identification or affiliation is believed to be acquired during social interaction and be transmitted from parent to child. This view of partisanship identifies party identification as the most stable affective attachment as well as the most powerful predictor of a person’s political attitudes and behaviors [@hopkins2009partisan;@keith1992myth;@miller1996new]. Given its stability, political identification serves as a potent cue in guiding the attitudes and behavior of the average person [@goren2005party;@dowding1983meaning]. Following this logic, I propose the hypothesis that individuals are motivated to share disinformation congenial to their party identification and political preferences. **H1: Partisans, including individuals who identify as Democrats and Republicans, are more likely to share partisan disinformation congenial to their political beliefs.** In other words, Democrats are more likely to share disinformation favorable towards Democrats and unfavorable towards Republicans, and vice versa. ## Political knowledge      In addition to party identification, one's level of political literacy and sophistication presumably contributes to disinformation sharing behavior. A central concept in the studies of public opinion and political behavior, political knowledge has been employed as an effective approach to measuring political sophistication. One of the fundamental assumptions in political science research is that it is crucial for an informed public to be equipped with a certain amount of political knowledge in order to engage in political participation. Defined as “the range of factual information about politics that is stored in long-term memory [@carpini1996americans,p.10]” political knowledge is usually measured with a battery of questions with respect to familiarity of current events and public policy [@carpini1993measuring]. Simply put, political knowledge is the key to information literacy in the era of information disorder.      Political knowledge affects civic participation both quantitatively and qualitatively. Citizens are expected to possess a certain amount of political knowledge to make informed and rational decisions about political issues and candidates. A wide array of scholarship has shown that individuals with higher levels of political knowledge are more likely to be politically active [@leeson2008media;@galston2001political;@tolbert2003enhancing;@carpini1993measuring]. In other words, they are more likely to cast their votes, to engage in political campaigns, and to make political donations [@carpini1996americans]. Lacking political knowledge, however, posts challenges for individuals to make sensible political decisions as they are unable to “integrate new information into an existing framework (built upon their political literacy) [@galston2001political, p.223]” to assist them assessing the current status quo of politics. If decisions were made for the sake of expediency rather than sensibility, it rarely would turn out to be the best decisions. To recapitulate, political knowledge translates to the abilities of reaching consistent and well-thought-out judgments in democratic governance.      In the low – information model put forward by @popkin1999political, low levels of information encumber public discussion, and thus lead to lower levels of acceptance of democratic values. As a result, individuals with lower levels of political information or literacy are less likely to devote to electoral politics. Even though another vein of scholarship suggests that individuals with lower levels of political knowledge may still be able to make reasonable judgments following cognitive shortcuts, heuristic devices, or psychological cues [@mcdermott1998race;@rahn1993role], it has been empirically established that political knowledge is an important factor in maintaining political belief and ideology consistency [@johann2012specific]. All told, citizens who are generally more informed about politics tend to engage more extensively and actively in democratic participation.      Indeed, in political communication research, several decades of empirical inquiry into citizens’ political knowledge or political literacy suggests that high levels of political knowledge serve as an indicator of adherence to democratic principles and active political participation, and acceptance and understanding of democratic values. Similarly, @lanoue1992one's account on informed citizenry offers a wealth of evidence that political knowledge fosters citizens’ acceptance of democratic values and judgments and enlightened self-interest; facilitate them to participate in civic and political life; help them express their high-quality opinions to shape the process of policy making.      However, this rich literature of political knowledge does not address the question of whether individuals with higher levels of political knowledge are less susceptible to false claims and political disinformation. Based on the discussion above, it is logical to assume that individuals with a higher level of political knowledge in a democratic society are less likely to share political disinformation. To further investigate this question, I propose the following hypothesis, **H2: Individuals with higher levels of political knowledge are less likely to share political disinformation on social media. **Put differently, one’s level of political knowledge is negatively correlated to his or her disinformation sharing behavior.      In addition to one's political identification and political literacy, this study also considers other potential predicted factors, including one's social media engagement, educational level,sex, age, and the information's credibility, that may contribute to disinformation sharing behavior. ## Social media engagement      Social media platforms, such as Facebook, Twitter, have become the fertile ground on which disinformation sown the seeds [@pennycook2019fighting]. Accessibility and affordability are the primary reasons why social media platforms may be especially conducive to disinformation. According to a 2019 Pew report [@shearer2019americans], social media has become a primary information source for news consumption as more than half of the U.S. adults often or sometimes rely on social media for news consumption. According to @allcott2017social, the accessibility of the social media supply chain has enhanced the spread of disinformation because the costs to produce and spread information on social is marginal. More importantly, the cost of spreading disinformation on social media platforms is much lower than that on traditional media platforms. As @allcott2017social note, “the fixed costs of entering the market and producing content are vanishingly small. This increases the relative profitability of the small-scale, short-term strategies often adopted by fake news producers, and reduces the relative importance of building a long-term reputation for quality (p. 221).”      Building upon the theoretical framework I sketched above, I propose the following hypothesis, **H3: The more engaged an individual is in social media, the more likely the individual shares political disinformation. ** ## Education      An individual’s level of education affects his or her ability to recognize disinformation [@khan2019recognise]. A recent study conducted by the University of Kansas surveyed how vulnerable populations shared COVID-19 pandemic related disinformation and found that one’s level of education is associated to abilities to identify such disinformation. Individuals with higher levels of education were more likely to recognize disinformation. However, there have been few research studies conducted on how one’s educational level shapes political disinformation behavior. Given that highly educated individuals are usually perceived to be more cautious about sharing unverified information, I hypothesize that education level is negatively correlated to one’s disinformation sharing behavior. **H4: Individual with higher levels of education are less likely to share political disinformation on social media than those with lower levels of education. ** ## Gender      The literature on gender in electoral politics suggests that men and women differ in political interests and participation. Prior research shows that men are usually more involved in political conversations and activities, and women tend to show less interest in social and political affairs and are less likely to hold political leadership positions [@conway2001women;@teele2018ties]. Studies further show that women lag behind in political activities and in shaping the democratic process [@kittilson2012gendered;@paxton2020women]. In addition, women score lower than men on political literacy and knowledge [@fraile2014women;@jerit2017revisiting;@lizotte2009explaining;@mcglone2006stereotype;@miller2019knows]. A recent study provides more empirical evidence to such conclusions, in which they found that men are indeed more likely to express interest in social and political issues. However, their research also points out that survey questions designed to measure political interests are usually "male-oriented." In other words, the questions used in political surveys to prompt respondents to report their political preferences tend to reflect political interests of men rather than those of women [@ferrin2020gender]. This finding offers an insightful implication: It might be too early to jump to the conclusion that men are simply more politically literate than women. There exists a possibility that men and women are simply have different political interests. Therefore, it requires further investigation to examine how sex contributes to predicting disinformation sharing behavior. To recapitulate, gender may be a factor influencing people’s understanding of and interest in politics. This study uses a non-issue-based survey to address the question of whether men are more likely than women to share disinformation. Given the extensive literature on gender politics, I hypothesize that men are more likely to spread disinformation on social media, hence, to re-frame, **H5: Male individuals are more likely than female individuals to share political disinformation on social media.** ## Age      It has been widely assumed that age itself is a predictor of disinformation sharing behaviors: The elderly appear to be more likely to spread political false information. Surprisingly, there has been relatively insufficient empirical evidence lending support to this claim. One of the very few studies with respect to age in disinformation and fake news susceptibility was conducted by @guess2019less. Situated in the context of Facebook, they investigated how individuals visited fake news domains and found that individuals aged 65 and older tended to share more fake news they read about on the fake news websites to their Facebook pages as they lacked skills and tools to determine the veracity of online news. Fake news articles from fake news domains, however, is not tantamount to political disinformation as disinformation is not necessarily generated by fake news websites. Thus, it remains unanswered whether older people are more likely to share political disinformation than younger age groups, although I would assume that age is positively associated with partisan disinformation sharing behavior. These considerations lead to the following hypothesis, **H6: Older people are more likely to share political disinformation on social media. ** ## Information credibility      Finally, numerous previous studies have investigated the psychological mechanisms of disinformation processing and shown that source credibility profoundly (re)structures individuals’ assessment of information veracity [@axsom1987audience;@baum2009shot;@berinsky2017rumors;@swire2017processing;@vraga2017using]. As these studies suggest, seeking heuristic cues to help evaluate veracity of information from political figures or other trusted sources is a rational use of individuals’ finite time and limited cognitive resources. In this regard, relying on politically polarizing information sources may result in their misconceptions of complex political topics such as policy or current affairs. The following hypothesis is therefore proposed to test if the credibility of information sources shapes an individual’s behavior of sharing disinformation. **H7: Individuals are more likely to share political disinformation on social media if the disinformation references credible sources.**