The effect of source cues on quality perception: an experimental
study investigating audience quality perception of Italian online newspapers
political content
Giacomo Galardini 10832750 Master’s Thesis
Graduate School of Communication Master Programme Communication Science
University of Amsterdam Supervisor: Dr. Jonas Lefevere
2 Abstract
This study investigates the effects that source cues have on perception of news quality in the online audience of Italian newspapers. Based on Gentzkow and Shapiro’s Model of Media Bias, the research explore the interaction between source with political ideology when evaluating political content from a certain news outlet. This study uses an online experiment ( N = 138) to examine how the source cue of the three most important Italian newspaper affects perception of news quality. Results show that respondents in the control condition had an higher quality perception than the ones in experimental conditions. Moreover, ideological distance did not show to have a significant interaction effect on perception of overall quality. Future lines of research and practical implications have also been discussed.
Keywords: Source cue, News Quality, Political Ideology, Media selectivity, Partisanship, Italy
3 Introduction and Research Question
A recent Pew Research Center study1 on political polarization and media habits of the American electorate showed that news sources play a crucial role in triggering trust or distrust in the news. This affects the perception of quality and reliability in the audience towards specific news outlets or media institutions. In the survey, which highlighted a significant correlation between respondents’ political inclination (or ideological consistency) and trust levels of news sources, liberals showed that they trusted 28 out of 36 different news outlets, but among the ones they distrusted (e.g. Fox News or the Rush Limbaugh Show) the shares of distrust were quite high (respectively 71% and 85%). In contrast, among conservatives, only 12 sources out of 36 were rewarded with higher share of trust rather than distrust.
This issue is only one of the most recent examples of the ongoing debate about the way in which partisan source affiliation, ideological selectivity, media polarization and fragmentation and more importantly, source cues affect news perception and quality evaluation (Gentzkow & Shapiro, 2014; Go, Jung and Wu, 2014; Iyengar & Hahn, 2009; Sundar, Knobloch-Westerwick & Hastall, 2007). A source cue is one of the most important heuristics that influence the way in which audiences select and perceive the quality of the news, guiding their judgment on the relevance, credibility and importance of a certain news story. A cue is defined by Sundar (2008) as a “nominal marker” (p.74) or anything in the context of media use that might serve as an indicator for the operation of a heuristic. A heuristic is simply an assessment rule, or a “mental shortcut” (Sundar, 2008, p.74) used in the audience for effortlessly estimate the quality or credibility of a given information. In the case of source cue, the focus is on the final agent that provides an information to the audience, e.g. the newspaper or media firm that published the news or, in the online environment, the
1 “Political Polarization & Media Habits in the American Public today” project survey, published
October 21th, 2014, accessed February 12th, 2015. Accessible on
http://www.journalism.org/2014/10/21/political-polarization-media-habits/#trust-and-distrust-liberals-trust-many-conservatives-trust-few
4 website from which the information was retrieved. Therefore source, along with other
interface cues or features like story placement, the presence/absence of a photo, and other editorial cues were studied in the literature of news quality perception (Goren, Federico & Kittilson, 2009; Graber, 1986).
Traditionally, a number of studies involving the concept of source cues in
communication science were referring to print newspapers (Rogers & Kincaid, 1981; Severin & Tankard, 1979). More recently, with the increasing importance of both the Internet and social media, we witnessed how digital media are becoming one of the main sources for news access around the world, especially for younger generations (Levine & Newman, 2014). Therefore, source as a medium for perceived quality and trust – due to the blurred internet environment - has become of interest to several scholars (Lucassen & Schraagen, 2011; Messing & Westwood, 2012; A. Williams, 2012), showing the increasing importance of online digital media in the news environment.
In addition, several studies on partisan source affiliation argue that people tend to ascribe certain partisan positions to media entities and evaluate content based on perceived individual ideological congruence (Bennett & Iyengar, 2008; Stroud, 2007). Moreover, studies reported how an increasing fragmentation of the media environment limits the
diversity of information available to the audience, polarizes individual-level attitudes (Stroud, 2010) and also increases ideological homogeneity regarding partisan selectivity (Levendusky, 2010). With regard to this, scholars discussed the impact that media bias has on news
perception on different levels. Most importantly, they argue that news quality is connected to outlet reputation, which has to match pre-existing individual evaluations of the outlet
(Gentzkow & Shapiro, 2006). Subsequently, they explored the effects that partisanship and ideological selectivity have on news information process (Tworzecki & Markowski, 2014).
5 Despite the existing literature on the topic, more research is needed to unveil with accuracy the relationship between source cues and news perception, more specifically quality evaluation. In other words; to what extent does the source cue, intended as the media firm that provides the news, affect the perception of new quality and credibility? Many of the aforementioned studies focus on US or UK media environments neglecting other Western democracies and countries. This ignores the fact that in other countries, media and political system peculiarities might play a crucial role as contextual third variables in assessing news perception (Hallin & Mancini, 2004).
Therefore, the external validity of previous findings needs to be replicated in different contexts, taking into account the perception of quality by the audience. In order to bridge this gap, this study aims to investigate the correlation between media source cues and political ideology on online quality news perception in Italy. This country offers a very particular and peculiar media and political environment, in which media pluralism is in danger by the lack of a mainstream separation between public and private media (Hibberd, 2007) and conflict of interests (Durante & Knight, 2012). An increasing perception of media bias (Ragnedda, 2014) and the historically polarized media environment (Mancini, 2013) are jeopardizing the
impartiality and balance of political news coverage. In this particular media landscape, the current study aims to bridge this gap by answering to the following research question:
RQ: To what extent does source cue of online newspapers in Italy affect readers’ news perception of quality?
Perceptions of news quality
This study extends Iyengar and Hahn (2009) work on ideological selectivity of media content to include analysis on the effect that source cues has on audience’s news perception.
6 The aim of this research is to assess whether perception of news quality can play a role along with political ideology when explaining the reason why audience select a certain news outlet over another. Moreover, this study represents an unprecedented effort to study the effect of source cues on quality perception in Italy, adding to a field that is mostly dominated by American studies. Extant scholarship has never considered that perception of quality might help to explain the rationale behind news selection. Therefore, I will provide some literature background by defining firstly my dependent variable, news quality perception, providing some conceptual definitions that serve to clearly narrow down the vast literature on different operationalizations and measures. Secondly, an overview of the different concepts of source cues and their underlying importance for studies in communication science will be presented. Finally, a clear definition of political ideology and its moderating role on news selection process will be presented.
In a world where rapid technological innovations are becoming increasingly important, the media is changing as fast as ever. Not all the changes are for the better, however. The consequences of media digitalization, such as cost and time constraints in online newsrooms, can lead to inaccuracy and misinformation (Craig, 2011). Competition on the web and general changes in media systems, such as increasing hyper-commercialization, market pressure and economic difficulties (Plasser, 2005) could force newspapers to decrease journalistic quality standards to remain competitive, and the increasing role of user generated content and citizen journalism in online media are threatening traditional journalistic norms (K. Williams, 2011). Therefore, if media users do not receive high-quality, relevant and accurate news, they cannot act as well-informed and active citizens in a democracy
(Habermas, 1984). In a recent study on the consequences of hard and soft news exposition among Dutch readers, it was found that exposure to low quality, soft news exposition was associated with higher levels of political cynicism (Boukes & Boomgaarden, 2014).
7 However, this argument is based on the assumption that the audience can assess fairly assess the quality of the news, neglecting the question whether recipients are fully aware of the normative quality criteria for information evaluation. Defining the quality of news, or
journalistic products in general, is a tricky task, that depends on the perspective from which it is viewed. Scientist defined news quality, on a normative level, as relating to the functions of news media in modern democratic societies (McQuail, 1992). Nevertheless, it is mostly agreed that there is no such thing as objective quality in a journalistic item itself, but only certain specific indicators that can be used to interpret and code high or low quality (Sundar, 1999). Normative news quality criteria are quite abstract and complex, and even if the audience has those criteria at hand, it lacks background knowledge and expertise of basic journalistic skills to judge news content accordingly (Urban & Schweiger, 2014).
Theoretically speaking, Urban and Schweiger (2014) highlighted six basic quality dimensions to assess news content: diversity, relevance, accuracy, comprehensibility, impartiality and ethical standards. However, in research about media quality, recipient’s quality evaluations are less appreciated than quality evaluations by scientists or media experts, because it is generally argued that the audience is unable to judge due to a lack of normative knowledge (Urban & Schweiger, 2014).
Nevertheless, some scholars (Sundar, 1999; Sundar & Nass, 2001; Sundar, 2008) argue that while researchers have always investigated recipients’ news perception through normative criteria, the audience itself might have different perspectives on the quality of news. In his exploratory research, Sundar (1999) attempted to investigate on a semantic level which labels and adjectives - on a list of 92 items - recipients would have used to describe the quality of certain news content, claiming that not enough attention was dedicated to the underlying psychological dimension that recipients use during evaluation. As a result, he pinpointed four criteria. First credibility, defined as the global evaluation of the objectivity of
8 a story. Then quality, which means the degree or level of overall excellence of a news story and the most important central (informational) factor (Sundar, 2001). Representativeness of a news story is defined as a summary judgment of the extent to which the story is representative of the category of news. Liking, finally, measures the overall affective reaction that
individuals feel towards a story.
Those concepts are therefore considered closer on a cognitive level to the factors and measures that the audience uses while assessing the quality of news items, rather than the normative ones. This assumption is made by the findings of Urban & Schweiger (2014), in which recipient found it hard to evaluate normative criteria of ethics, objectivity and
comprehensibility of a news item, and relied more on media brand images, which conveyed an important heuristic when evaluating news quality. If the audience has to choose the
preferred media outlet for news consumption, it will evaluate also more cognitive criteria such as quality and the mere liking a news outlet rather than another one.
For this reason, the current study will use those four categories when measuring news quality perception. Perceptions of news quality are inherently important, as they determine whether and how audiences accept the information contained in the news coverage. This thesis assesses the impact of source cues on perceptions of news quality. However, news quality perceptions are not only influenced by the content of the news itself. If the provider or the origin of news content has proven to have an influence on audience capability of quality judgments, it is important to accurately determine to which extent and why this happens. In the next chapters, the concept of source cues will be developed first, then this will be related to political ideology, drawing hypotheses for this study.
Source cue: a definition
The research on effects of source cues in print and television media started quite early in time, but had to deal with some confusion over a clear distinction between source, medium
9 and author. In fact, Hovland and Weiss (1951) with their first study on this topic, confounded source with media channel by comparing major publications with famous individuals on a credibility dimension. Therefore, in order to make a clear definition for the concept of source, I use Sundar’s (2001) three different conceptions of communication sources, namely
technological source (the medium that conveys the information), receiver sources (the individual who selects, filters and recommend content), and visible sources. For the purpose of the study, visible sources cues are defined as “the sources seen by the receiver to be delivering the message or content”(Sundar, 2001, p. 58) where “gatekeepers” or media corporations labels are the first examples of visible sources. In the case of online newspapers, “source can be identified with either the reporter (in the form of a byline) or the collective editing staff of the newspaper (in the form of the masthead)” (Sundar, 2001, p.58). Therefore, I adopt this definition of source cue in order to investigate the impact of these types of visible sources on news quality perceptions.
Based on this source distinction, various studies (Messing & Westwood, 2012; Metzger & Flanagin, 2013; Sundar et al., 2007) have investigated the role of sources in news evaluation, providing evidence and a theoretical rationale to explain their relationship. One important dimension of this interaction is to investigate whether audience can actually distinguish between news source and news content when attributing evaluations of quality.
The same problem occurs especially with the credibility dimension, since the audience, especially with online news, transfers his credibility assessment from the source cited in the news article to the website itself (Metzger & Flanagin, 2013). In fact, the idea that individuals transfer their perceived credibility of the source of information to the quality of the information itself is posited in the 3S-model (Lucassen & Schraagen, 2011). The model posits that trust judgments and quality perceptions depend on three features of the
10 model, based on cognitive heuristic, was put forward as a strategy to consider the source of information (Kang, Bae, Zhang and Sundar, 2011), which suggests that “earlier interactions with a particular source may serve as a cue for the credibility of the current information” (Lucassen & Schraagen, 2012, p. 568), especially when receivers lack in skills to verify the accuracy of information.
Therefore, if a user has a positive experience with information from a particular news organization, he can decide to trust new information from that source without constantly and singularly evaluating the quality of news content, especially in an online context in which there is a low motivation and engagement with the text, as suggested by the Elaboration Likelihood Model (ELM) (Massaro, Petty & Cacioppo, 1988). The exact opposite can also occur: the less we are acquainted with a source, especially when we do not have the chance to verify the information at hand, the more we will engage systematically into assessing the quality of that news.
The evaluation occurs not only at a semantic level of text content, but also takes into account the surface features, such as references, pictures, layout and structure of the text. However, source cue has been proved by the model to be the most important factor for readers when assessing credibility of a text. In their study, Lucassen and Schraagen (2011) showed when readers lacked in expertise in evaluating the accuracy of Wikipedia articles, they still considered them to be high in quality and credibility. Hence, for the purpose of this study, in which source cues play an important role in the evaluation of credibility and quality of online news, the 3s-model presents a more interesting approach and gives a more accurate and detailed insight to the research. All in all, a source cue provides the user with the
knowledge of the origin of the information, and this mere fact enhances trust, contributing to increase the overall perception of credibility (and therefore quality) of the information itself (Lucassen & Schraagen, 2012).
11 Moreover, Koh and Sundar (2010) suggest that when choosing between sources, respondents are more likely to believe a recognizable source that is judged as more credible compared to an unfamiliar source, without carefully examining the actual content or source credentials. In online news quality evaluation studies, the names and the logo of news agencies, which in experiments are used as the main interface cues of online news websites, are generally closely related to the heuristic of news content quality; thus, the names of news agencies have been widely used as manipulations of source cues (Go et al., 2014, Kang et al., 2011).
Therefore, based on the 3s-model and on Sundar’s definition of source, I expect that the presence of source cues will lead to a more positive evaluation of the quality of the news article, as proposed in the following first hypothesis:
H1: The presence of the online newspaper source cue will lead to a perceived higher quality of the news content of the article, while the absence of the same cue will lead to a perceived lower quality of the same news content.
Political Ideology and its effect on news perception
According to a vast body of political and communication science studies, media effects are not the same for everyone, i.e. they are contingent on so-called moderators, which determine if an effect is stronger or weaker for a certain individual and/or in a certain context (de Vreese & Lecheler, 2012; Schuck & Feinholdt, 2015). One of the most prominent and important variables studied in relation to the news environment is individual political ideology, defined as a “set of beliefs about the proper order of society and how it can be achieved”, (Erikson & Tedin, 2003, p.64). Political ideology affects news perceptions and moderates news assimilation (Tworzecki & Markowski, 2014).
Existing research on individual-level political ideology investigates the interaction with the information environment (print, online and television media) in different directions.
12 For example, a study on selective exposure (e.g. the ability to select the news sources that the audience prefers) in political campaign communication showed that the degree of ideological partisanship has a moderating role on audience’s news selection, enhancing anticipated agreement (Iyengar, Hahn, Krosnick and Walker, 2008).
Another example that underscores the importance of audience’s prior beliefs in relation to assessments of quality and credibility of news items is the model of media bias (Gentzkow & Shapiro, 2006). According to this model – which is based on economic models of bias theory - a given media firm which aspires to establish itself as a reliable and successful company, will have to provide accurate information. However, if the quality and reliability of this information is hard be checked directly, consumers will have to rely on past reports, which the media itself will tend to distort according to consumers’ prior belief. In other words, media firms would economically benefit from the alignment to given political slants that go in the direction of their audience prior beliefs. The model provides evidences to support the influence of individuals’ prior beliefs on quality assessments. Moreover, it suggests a correlation between the editorial position of newspapers and the views of their readers. However, the assumption that media firms would abide to individual beliefs altering the facts probably undermines the ability of the audience to discern the actual quality and trustworthiness of the news, which was not measured in the study.
Depending on individual pre-dispositions or characteristics (in the case of this study, political preferences), respondents might ignore or rate more negatively sources that do not match their political predispositions, but also perceive information of an ideologically consistent source as more trustworthy compared to the same content conveyed by another competing source (Iyengar & Hahn, 2009). In their study Iyengar and Hahn showed that Republicans (party ideology) and conservatives (political ideology) preferred Fox News while more liberals and Democrats selected CNN and NPR over other media sources, concluding
13 that as the audience becomes increasingly polarized over political news, media owners will follow the trend of increasingly biasing news content, contributing to media fragmentation and attitude polarization in the national news environment.
Moreover, studies on political ideology and online news perception (Gentzkow & Shapiro, 2015; Messing & Westwood, 2012) argue that in the absence of any other
meaningful selection criteria (besides a title), people would rely on attributes of the source about their decision, “including the extent to which that source is likely to provide content that is ideologically consistent with their prior attitudes” (Messing & Westwood, 2012, p. 6).
However, these studies are either focused on specific conceptualization, cognitive and behavioral processes of news assimilation, or methodological implications (Murphy &
Westbury, 2013), ignoring the perception of quality in political news, which is not only crucial to determine the success of media firms in the competitive market (Sundar, 2001), but also to keep high standards for quality journalism (O’Sullivan & Heinonen, 2008). Therefore this study aims to implement the literature on interaction of political ideology and news information environment by investigating the effect of the former on news perception. I believe that political ideology and anticipated agreement would carry with themselves a higher perception of quality and credibility of the audience preferred news content. Just the mere fact that a news item is provided by an ideologically consistent source is not enough to justify news selection behavior: users must also like what they read. Hence, the following hypothesis is formulated:
H2: The smaller the ideological distance gap between respondents’ political ideology and newspaper perceived political slant, the higher the overall perceived quality of the online news will be and vice versa.
Methods
14 This study was designed as a 4 ( 3 experimental groups with source cue manipulations: center-left (La Repubblica), center (Il Corriere della Sera) and center right (Il Giornale), plus a control group without visible source cues) group between-subject online experiment. An experimental research design was preferred over a quasi-experimental because the use of a control group and the randomization of the manipulation would increase the internal validity of this study ensuring causal relationship between the variables (Halperin & Heat, 2012). A quantitative online survey questionnaire, which was available for over ten days2, mostly featured closed ended questions. Although a recent study has showed that using a particular computer software for measuring news perception (Iyengar et al., 2008) would guarantee a more accurate and reliable respondents’ answer, an online survey experiment was the best solution given costs and time constraints. Moreover, this method has been used by a large majority of previous studies (Go et al., 2014; Kang et al, 2011; Lucassen & Schraagen 2011; Urban & Schweiger, 2014).
Method overview
All participants were randomly assigned to one of the four experimental conditions. The source cue was manipulated by altering the name of the online newspapers. Is important here to give some details about the rationale behind the choice of Italian online newspapers as the context for the experiment. The choice of online news articles as the unit of analysis is due to increasing importance of online and digital media in the Italian audience, a trend that has been increasing in the last years. In fact, in the last 2014 Reuters Digital News Report3, Italian
2 From 18th to 29th May 2015. 3
Data retrieved from the 2014 Reuter Institute Digital News Report, edited by Nic Newman and David A. L. Levine, accessed on 12th February 2015 on
http://reutersinstitute.politics.ox.ac.uk/sites/default/files/Reuters%20Institute%20Digital%20News%20Report%2 02014.pdf
15 online newspapers were more often used (67%) to access the news in comparison to their print counterparts (55 %). Since the aim of this paper is to test whether the findings of source cues and political ideology in the American literature can be observed also in the Italian media landscape and on the Italian audience, a few words on differences and similarities between the two media systems and journalistic traditions are needed. In a recent comparative analysis of political coverage between several European and American high-quality
newspapers, Esser and Umbricht (2013), found that in comparison to the American one (which favors objective and interpretative journalism), the Italian political coverage, being the country representative of the Polarized Mediterranean media system (Hallis & Mancini, 2004), publish less “straight news” and more opinionated items and present a less objective coverage and a more negative and conflict oriented articles.
Regarding public opinion and information environment, the Italian public sphere does not differ substantially from the portrait given by Iyengar and Hahn (2009) of the American one: Mancini (2013) reports the high levels of polarization of the media and segmentation of the audience also in the Italian public sphere. The only notable difference is perhaps the arrival of Silvio Berlusconi (the controversial former Prime Minister and media tycoon) on the national media business stage, who incremented the already existing “rooted partisan character of the Italian news media” (Mancini, 2013, p. 338) and partisanship of news content (Durante & Knight, 2012), also increasing the division of the audience along the lines of polarized political attitudes (Mancini, 2000). In fact, the conflict of interest that Berlusconi established in the Italian media system (being himself owner of Mediaset, Italy’s major commercial television and Mondadori, the biggest publishing media corporation) poses a threat to media pluralism in the southern European country. While in office, he wielded a high degree of political influence on the public service broadcasting company (Hibberd, 2007).
16 However, one of the limitations of these studies on the Italian media system situation is that they have been limited in their analysis to mostly TV and broadcasting channels, ignoring the print but particularly the online news media websites.
In addition, regarding news perception, Italy ranks first in the 2014 IpsosMori4 Index of Ignorance, stressing the problematic nature of Italian audience’s news perceptions (and misperceptions). According to this report, the Italian audience over estimated the salience and gravity of certain issues portrayed by the news (e.g., stating that 30% of the population in Italy consists in immigrants, against the 7% of reality). Finally, regarding media system and environment, Italy ranks 73rd in the 2015 World Press Freedom Index5, one of the 5 worst positions of EU countries.
Therefore, in order to replicate and test whether this peculiar contextual factors could affect the outcome of this study, three online versions of well-established Italian print
newspapers were used, which widely represent the different political inclinations of Italian political spectrum: La Repubblica, Il Corriere della Sera and Il Giornale. La Repubblica, founded in 1976 and owned by the L’Espresso editorial group, which is controlled by De Benedetti, a businessman who has interests in various sectors. La Repubblica ranks first in the 2014 Reuters Institute News Top Brand chart for weekly usage and historically has always been linked with center-left political attitude. Il Corriere was founded in 1876 and is owned by RCS Mediagroup ( whose largest shareholder is FIAT, Italian biggest automaker
company). The newspaper has always been traditionally conservative without ever leaning decisively towards left or right wing positions, and ranks fifth in the same 2014 Reuters chart.
4“Perception are not reality: things that world gets wrong”, IpsosMori Global Survey, published
October 29th, 2014, accessed February 10th, 2015. Accessible on
https://www.ipsos- mori.com/researchpublications/researcharchive/3466/Perceptions-are-not-reality-Things-the-world-gets-wrong.aspx
5 World Press Freedom Index 2015, Reporters Without Borders For Freedom of Information, published
12th, February 2015 and accessed 16th February 2015 on: http://en.rsf.org/world-press-freedom-index-2015-12-02-2015,47573.html
17 Finally Il Giornale, owned by Silvio Berlusconi, the Italian former prime minister and current leader of the right wing party Forza Italia since 1996, is the most favored newspaper for the right-wing electorate (Mancini, 2000).
Participants
A total of 138 respondents participated in the experiment, although only 102
completed the online questionnaire. Item non-response caused a fluctuation of the N number, but due to the small N I preferred to keep respondents who did not answer all questions. If only respondents who never answered “don’t know” would have been used, the N would become too small to make any meaningful inferences.
The sample was collected on a voluntary base by posting an ad on the website jobplacement.unipi.it, which is the most used for students and post-graduate to find a future working occupation or academic networking career of the University of Pisa, in Italy. Clearly there are limitations in using students sample (Sears, 1986) but Druckman notes that “the behavior of student participants does not significantly differ from the behavior of non-student participants” (2001, p. 1046), and that there is reason to assume the generalizability of such a sample. Moreover, participants selection was carried out through convenience and snowball sampling. I made use of personal social media and email contacts to diversify the age range of respondents, which average was 28 (SD = 6.40) with a slight majority of males (51%) over females. Regarding their level of education, 30 % had at least an high school diploma, 32 % a bachelor degree and 23 % had completed their master studies.
Procedure
All participants received a link that directed them to the online experiment. The online survey was conducted on Qualtrics, an online private research software, which allows
18 researchers to insert a news story, a multimedia article or any other stimuli in the online questionnaire. Before starting the survey, an explanation of the task was provided.
Participants were informed that they had to evaluate the perceived quality of a political news story, without specifying how to perform this task. After reading the instructions, respondents decided to participate by clicking on the ‘next’ button. After doing this, they were asked to provide information about their sociodemographics first. Subsequently, all participants were asked to read a news story on a simulated online news page (either La Repubblica, Il Corriere della Sera, Il Giornale or without any specific source cue, but with the exact same layout of the respective media corporation websites). After having spent as much time as they wanted reading the news story, but at least 45 seconds, in order to avoid skipping the task,
participants moved to the next page to complete the main questionnaire about the perceived credibility, quality, and liking of the news story. After answering the main questionnaire, participants completed a manipulation check questionnaire, placed in the end to avoid priming effects. Upon completion of the survey respondents were informed about the true purpose of the study.
Stimuli
In this study, a real news story written by ANSA, the most important Italian wire service News Agency was used. The content of the article was political, but it dealt only with the biography of the recently elected President of the Republic Sergio Mattarella, the highest figure in the national institutions. The choice of a biography is due to the fact that, while still dealing with a political article, it grants a certain level of objectivity and reliability of the information conveyed in the article (provided by the political neutrality of the News Agency), which is necessary to secure that the news source will be the only possible heuristic cue or mental shortcut that respondents will use for their evaluation (e.g. see Valentino, Beckmann
19 and Buhr, 2010). Moreover, a biography has the least interpretable political cues in terms of content. There was no author judgment or evaluation on Mattarellas’ life, no personal opinion and no omissions. Furthermore, Mattarella is relatively unknown to the public (was elected in February 2015) and therefore cannot be linked or framed in a specific political orientation or party inclination. To make the layout of the simulated webpage, screenshots of the existing three news websites directly were captured, and then original articles on both screenshots were replaced with the chosen article. Thus, original banners, advertisements, news menu bars, and more importantly, news agency logos were retained unaltered.
Measures
News Perception
Perceptions of the news article were measured in terms of participants’ evaluations on the content of three broad categories (credibility, liking and quality) used by previous studies as the criteria for the evaluation of perceived quality of news content (Go et al., 2014; Sundar, 1999; Sundar & Nass, 2001). These measurement were adapted firstly by deleting the
category of representativeness. In fact, Sundar (1999) defines it as a “summary judgment of the extent to which the story is representative of the category of news” (p.381). Since my stimulus consisted in a politician biography, I decided that it was already enough
representative of traditional political coverage. Secondly, I merged some of the adjectives used to describe the perceived quality in order to facilitate the respondent task and also to avoid acquiescence response bias (Leighley, 2010). The detailed measures will be as follows.
Perceived credibility: perceived credibility of the news story was assessed on a 7-point likert scale (ranging from ‘‘strongly disagree’’ to ‘‘strongly agree’’) and measured with the question: “on a scale between 1 and 7, where 1 represent strongly disagree and 7 strongly
20 agree, how much do you agree that the content of the article you just read can be considered as” using 4 adjectives (i.e., believable, accurate, biased, objective) (α = 66).
Perceived quality: to measure perceived quality, 5 adjectives (i.e., clear, coherent, important, concise, well-written) were used (α = 75).
Perceived liking: perceived liking was assessed on a 7-point scale (anchored between ‘‘strongly disagree’’ and ‘‘strongly agree’’) using 4 adjectives (i.e., boring, , interesting, enjoyable, informative) (α = 75).
Overall Quality Scale: The Overall Quality Scale was obtained by merging the results of the individual sub scales of credibility, quality and liking, in order to have a more useful and practical measure to test the complete perception of quality (α = 75).
Ideological distance respondent-source
In order to answer the research question and hypotheses, I created a specific variable, which I named Ideological Distance Respondent-Source. This scale reports the absolute distance between the respondent’s own Political Orientation, measured with a self-assessment question on a scale from 1 (left) to 5 (right), and the perception of the political orientation of the source that respondents’ got as a stimulus in the experimental conditions, ranging from 1 (left) to 5 (right). As a second step, I divided the scale in three different sub groups that I labeled respectively as Short or No Distance (N = 27) which range from 0 to 1 in the scale, corresponding to the 35 % of the total (N = 78) of respondents exposed to source cues manipulation, Medium Distance (N = 36), ranging from 1,5 to 3,5 and representing 46 % of total respondents, and finally Large Distance (N = 15), ranging from 4 to 5 and counting as 19 % of the total population of valid response.
21 Results
Two questions were asked to assess whether the manipulation had succeeded. Firstly, participants were asked in which region President Mattarella was born. Secondly the
participants were asked from which newspaper website the news article was retrieved, in order to check if the source cue manipulation was correctly perceived. The manipulation check showed successful manipulation in both regards. Almost all the respondents (N = 99) across the four condition (N = 102) reported correctly about Mattarellas’ birthplace. For the source cue manipulation check, 74 out of 102 respondents correctly identified the source, as indicated by a chi-square test (χ2
(9, N = 102 ) = 130.99, p = .00 . Therefore, all manipulations
were successful.
The participants were randomly distributed over the conditions, with the N per condition fluctuating between 24 for Il Giornale and 28 for Il Corriere condition. The
experimental groups did not differ significantly to each other with respect to variables like sex [F(3,100) = 0.65, p = 0.58] and age [F(3,100) = 2.18, p = 0.09], with age being slightly higher for the control condition but still insignificant.
My first hypothesis (H1) was that the presence of the online newspapers source cues would trigger an higher overall perception of news quality, while their absence would have caused a lower overall news quality perception. I conducted a one-way ANOVA in order to compare the means of the overall quality perception of the news article across the three experimental conditions and the control condition. However, perception of Overall Quality did not differ significantly across conditions ( F (3,98) = 1.03, p = 0.38). In other words, source cues did not result in a higher overall quality perception. The results showed surprisingly that all three source cue conditions had a lower mean of overall quality in comparison to the control condition (N = 26), which scored higher (M = 4.53, SD = 0.95) in comparison to La Repubblica (N = 26), who placed itself second (M = 4.46, SD = 0.85), in
22 front of Il Corriere (N = 28),(M = 4.43, SD = 0.98) and Il Giornale (N = 22), scoring lowest (M = 4.10, SD = 0.82). Subsequently, I conducted a sub group analysis comparing each of the three categories - credibility, quality and liking - of the overall quality perception across the four conditions in another one-way ANOVA.
Although the stimulus did not show any significant difference among credibility (F (3,99) = 1.09, p = 0.36), Quality (F (3,99) = 1.27, p = 0.29) and liking (F (3,98) = 0.68, p = 0.57), the means of each category show some interesting patterns nonetheless. While the control condition showed the highest score for quality (M = 4.78, SD = 1.13) in comparison to the other experimental conditions, source cues conditions were overall perceived more
credible than the control one, with Il Corriere (M = 4.50, SD = 0.84), La Repubblica (M = 4.33, SD = 0.90) and Il Giornale (M = 4.21, SD = 0.72) all scoring higher than the control condition ( M = 4.12, SD = 0.74). Regarding the liking category, Il Giornale condition is the one which has the lowest score (M = 3.90, SD = 0.62) compared to both the experimental La Repubblica (M = 4.18, SD = 0.79) and Il Corriere (M = 4.18, SD, 0.86) conditions, and the control one (M = 4.18, SD = 0.85). The source-less condition scored higher in quality despite the absence of any surface features and regardless heuristics expectations, while the presence of media labels seems to have triggered an higher credibility perception, in line with previous researches. In addition, it appears that Il Giornale is the newspaper that respondents liked the least.
Overall, hypothesis H1 is contradicted by results, and has to be rejected. Among the sub categories, credibility is the only one that goes in the direction expected by the
hypothesis, although the difference is statistically insignificant: respondents in the source cues conditions considered the stimuli more credible than the one in the control condition.
My second hypothesis (H2) was that the smaller the ideological distance between respondents’ own political orientation and the perceived position of the source in the political
23 spectrum, the higher the perceived quality of the news will be and vice versa. In other words, I expected that the perception of Overall Quality of the online news article would be
moderated by the distance between the respondent’s and the source’s political orientation. In order to test H2, a one-way ANOVA was used to determine whether significant differences between the subgroups of distance existed in terms of overall quality perception. The results show that there is a marginally statistically significant effect (F (2,73) = 2.25, p = 0.11) on the Overall Quality scale which goes in the direction of my hypothesis. Graph 1 contains the mean plot scores for each of the three distance sub groups along the Overall Quality scale, which illustrates a decreasing tendency in quality perception scores, meaning that the higher the ideological distance between respondents and source, the lower scores for quality will result.
Graph 1: Means plot scores.
Subsequently, a two-way ANOVA was carried out to test whether the ideological distance respondent-source and the exposure to the three experimental conditions or their interaction had a statistically significant effect on the Overall Quality scale. In fact, the sub groups of Short/No Distance has the highest score for Overall Quality (M = 4.49, SE = 0.25),
4,49 4,4 3,73 0 1 2 3 4 5 6 7
Short/no Distance Medium Distance Large Distance
O ve ra ll Q u al ity Sc al e
Ideological Distance Respondent-Source
Mean plots of quality perception across
distance groups
24 followed by medium distance (M = 4.40, SE = 0.15) and large distance (M = 3.73, SE = 0.38). In other words, the bigger is the gap in political orientation between respondent’s own and source cues perception, the lower score the article will receive in the Overall Quality scale. The test of between subjects effects showed no significant interaction between ideological distance and exposure to condition in Overall Quality (F (4,67) = 0.608, p = 0.66).
Nevertheless, the multiple comparison in table 1 shows that both La Repubblica and Il Giornale show a decreasing direction of quality scores along a increasing one in ideological distance, while Il Corriere has a similar mean score across the sub groups, meaning that despite ideological distance, his quality mean scores remained unaltered.
It is nonetheless important to state that given the low N of some sub groups - resulted by the combination of conditions and the sub groups of Ideological Distance Respondent-Source scale - the interpretation of the results of the table necessitates caution. Especially for the Large Distance groups in La Repubblica ( N = 1) and Il Corriere ( N = 2) conditions, and for short/no distance group in Il Giornale condition (N = 1). The size of the respective sub groups across conditions shows that regarding political ideology, the sample (N = 102) was skewed towards centre left position (M = 1.73, SD = 0.94) on a scale from 1 (left) to 5 (right), with 84 % of the respondents scoring between left and center and only 16 % being center right or right wing oriented. This factor might explain the heterogeneity of the size of the sub groups.
Table 1: Overall Quality Perception by Conditions and Ideological distance respondents-source (Short/no distance, medium and large) for Hypothesis 2.
Ideological distance respondents-source
Short/no distance Medium distance Large distance
La Repubblica condition 4.55 (0.90) N = 15 4.48 (0.70) N = 10 2.92 (/) N = 1
Il Corriere condition 4.34 (1.20) N = 10 4.49 (0.83) N = 16 4.42 (1.58) N = 2
Il Giornale condition 4.58 (1.36) N = 2 4.24 (0.72) N = 10 3.86 (0.85) N = 10
Higher means signify higher scores on the quality perception of the news article (1-7 scale). Standard deviations appear in parenthesis next to means. N indicates the number of respondents per condition and sub groups.
25 Subsequently, a series of two-way ANOVA between Ideological Distance Respondents-Source, conditions and the three sub groups of Overall Quality perception scale (credibility, quality and liking) were conducted to see the direction of interactions across the Ideological Distance Respondents-Source scale in the specific subgroups of the dependent variable. For the subgroup of quality, results show no significant differences (F (4,68) = 0.59, p = 0.67) of the two independent variables. Here, as table 2 shows, subgroup quality highlights the same tendencies of the overall quality scale, with Il Giornale and La Repubblica having decreasing score along the ideological distance scale and Il Corriere showing a slightly higher score for medium distance ( M = 4.65, SD = 0.97) than short/no distance ( M = 4.58, SD = 1.19).
Table 2: Quality Perception scores by Conditions and Ideological distance respondents-source (Short/no distance, medium and large) for Hypothesis 2.
Ideological distance respondents-source
Short/no distance Medium Large
La Repubblica condition 4.65 (1.03) N = 15 4.64 (0.75) N = 10 2.80 (/) N = 1
Il Corriere condition 4.58 (1.19) N = 10 4.65 (0.97) N = 16 4.0 (0.97) N = 2
Il Giornale condition 4.6 (1.70) N = 2 4.26 (0.93) N = 10 4.09 (0.97) N = 11
Higher means signify higher scores on the Quality perception of the news article (1-7 scale). Standard deviations appear in parenthesis next to means. N indicates the number of respondents per condition and sub groups.
By looking at the credibility sub group, we can observe an marginally significant
difference between the conditions ( F (4,68) = 2.00, p = 0.10). In fact, table 3 shows that the Il Corriere cue is the one that was perceived more credible regarding the ideological distance, actually having the opposite tendency, with mean scores scoring lower for Short/No Distance ( M = 4.45, SD = 1.065), slightly increasing along Medium ( M = 4.48, SD = 0.76) and scoring higher for Long Distance ( M = 4.87, SD = 0.84), while Il Giornale condition dropping from higher score in the Short/No Distance ( M = 4.62, SD = 1.24) to considerably lower ones for the Large Distance ( M = 3.90, SD = 0.60). The same pattern emerges for La Repubblica, going higher on Short/No Distance mean scores (M = 4.45. SD = 0.82) and decreasing along when the ideological distance gets larger.
26
Table 3: Credibility Perception scores by Conditions and Ideological distance respondents-source (Short/no distance, medium and large) for Hypothesis 2.
Ideological distance respondents-source
Short/no distance Medium Large
La Repubblica condition 4.45 (0.82) N = 15 4.37 (0.78) N = 10 2 (/) N = 1
Il Corriere condition 4.45 (1.06) N = 10 4.48 (0.76) N = 16 4.87 (0.18) N = 2
Il Giornale condition 4.62 (1.24) N = 2 4.45 (0.70) N = 10 3.90 (0.60) N = 11
Higher means signify higher scores on the Credibility perception of the news article (1-7 scale). Standard deviations appear in parenthesis next to means. N indicates the number of respondents per condition and sub groups.
Finally, the liking sub group also showed a certain degree of marginal statistical significance (F (4,67) = 1.91, p = 0.12), where the medium distance groups, for a all conditions, showed a higher mean scores in comparison to the Short/No Distance one.
Therefore, respondents’ were not affected by ideological distance when assessing the liking of the article at stakes.
In conclusion, given the absence of significant results, hypothesis 2 has to be rejected. Nevertheless, the direction of the results of credibility and Overall Quality - especially for Il Giornale and La Repubblica conditions – show that higher ideological distance scores equal lower quality mean scores, while Il Corriere della Sera – while still being marginally
significant - seemed to be unaffected by ideological distance between respondents and source regarding the subgroup of credibility.
Discussion
The aim of this thesis was to investigate to what extent source cues affected readers’ perception of news quality of Italian online newspapers. I used theoretical discussion to detail Sundar’s conceptualization of source cues, the 3-s model (Lucassen & Schraagen, 2011) and the model of media bias (Gentzkow & Shapiro, 2006). In the results, it can be observed that the presence of source cues, contrary to previous research, is not able to trigger higher
27 perception of quality of news content by itself. As for the second hypothesis, ideological distance did not show to have a significant interaction effect on perception of quality.
Relatively to the effect of source cue on perceived quality, the first interesting result is that the control condition was perceived higher in quality than the newspaper counterparts. This result seemed to be controversial and unreasonable in the beginning. A more thorough examination lead to a couple of interesting suggestions. Firstly, it can be partly explained by the 3-s model, which states that if the reader has enough motivation to engage critically with the content of the news, he will cognitively process and assess the content in a more systemic way. Thus, since the article provided no other verifiable surface or source cue other than the semantic (content) cue, respondents attentively evaluated the central argument of the article, which was provided by a news agency and therefore met the highest journalistic professional standards of quality. Consequentially, respondents in source cues conditions might have been cued by source labels and avoided to critically assess news content, giving an average rating.
However, is important to highlight again that a higher quality perception for the control condition goes against some previous findings (Koh & Sundar, 2010): in fact the mere presence of a source cue was not found to “increase the appeal of news stories across all dimensions” (Iyengar & Han, 2009, p. 26), as consumers did not ignore anonymous news reports. In other words, source cues are not a guarantee for higher quality perception.
Perhaps the most interesting result is that source cues had a higher perception in credibility. Once again, the versatility of the 3-s model can provide a possible explanation: users might have passively relied on previous experiences with the online newspaper to assess their trust (Lucassen & Schraagen, 2001). The question remains why this happened only with credibility, and not with quality and liking. A possible explanation for this distinction might come from the media literacy of the audience (Vraga & Tully, 2015) which is the ability to consider the editorial staff of a newspaper as professionally equipped to judge, through
28 gatekeeping, whether an article can be considered credible or not. Systemic processing is time consuming and all in all, trust assessments are by nature heuristic to a certain degree. To corroborate this explanation, we can observe Il Giornale’s results on credibility. Berlusconi’s newspaper was perceived more credible than the control condition, even if it was the one with the smallest score on liking in comparison to the other two newspapers. Moreover, Il Giornale present some differences with the other newspapers in regard to surface cues: the layout contains less picture and a more opinionated, right-wing oriented selection of related articles and opinions about Italian politicians, like Mattarella, that might have influenced the
perception of liking of the page structure.
However, motivation or expertise regarding the stimulus topic was not measured, therefore no definite assumption can be made about the ability of the 3-s model to explain this results. Still, it can be reasonably assumed that audience had little knowledge about Mattarella since he is relatively unknown in the political arena, especially to younger generations.
Nonetheless, the fact that credibility is the only sub group that goes in the direction of the hypothesis is consistent with other findings (Kang et al., 2011), since credibility “has long been perceived as an attribute of the source” (p. 721), and trust in information was proven to be influenced by trust in its source (Lucassen & Schraagen, 2012).
The second hypothesis regarded the possible moderating effect that ideological distance between respondents’ own political orientation and the perceived position of the source could have had on quality perception. This hypothesis received only partial support. It is reasonable to assume that one of the reasons for this is the small N: a bigger number of respondents could have brought more statistically relevant results. Nonetheless, while we cannot report a moderating effect of ideological distance on perceived quality, we can state with more confidence that the results go in the expected direction of previous research (Gentzkow & Shapiro, 2015; Messing & Westwood, 2012) except for credibility, an
29 important exception. In fact, if we watch at the mean score for overall and sub scales, Il
Corriere is the one least affected by variation across distance in political ideology. While for Il Giornale and La Repubblica the audience was more keen on decreasing their evaluation score the more the distance increased, the same did not happen for Il Corriere. Especially with credibility, the mean score surprisingly rose along ideological distance. This means that if a newspaper is perceived as credible, audience would tend to consider it trustworthy regardless of its political slants. Thus, while it has been proven that anticipated agreement might be triggered by the ideological consistency (Jerit & Barabas, 2012), the reputation of a news corporation seems to be more relevant when it comes to news perception. However, the external validity of the study’s findings is undermined by the small size of Ideological
Distance sub groups. Therefore, a generalization based on these results, especially for La Repubblica and Il Giornale, needs to be drawn with caution.
The results of Il Corriere are the only one that go against the model of media bias proposed by Gentzkow and Shapiro (2006), showing that providing a perceived credible and quality news content might work as a bulletproof vest against ideological selectivity, source affiliation and polarization. This means that maintaining a higher perception of credibility is important to preserve the standards of journalistic work and attract an audience without necessarily being forced by the market logic to feed the audience with politically biased content. It is important to say thought that while it appears that the audience has higher rates of credibility for news websites with balanced political coverage, it is still unclear whether this feature is considered to be the most important factor that drives news selection.
To sum up, in this research the perception of quality seems to be affected by source cues to a certain extent. When source labels are the only distinguish visible features for assessing news content, their presence trigger a higher quality but a smaller credibility score. When political ideology come into play, quality and liking perception show a tendency to be
30 affected by ideological distance, while credibility perception seems to remain unaltered. The systemic differences between the American and the Italian models of political coverage (Esser & Umbricht, 2013) and the Italian tendency to polarization might explain these differences. However the abovementioned studies never took perception of news quality and credibility into account. Moreover, the two systems did not differ substantially in terms of media
fragmentation and ideological selectivity, but the fact that credibility stood out in the markers for news perception might also be explained by the fact that the Italian audience is used to evaluate credibility in a thorough manner. The media concentration and the influence that the political figure of Berlusconi plays in the Italian daily life might have affected this outcome. Further research should also take other news subjects and topic - rather than only political news – in order to measure the truthfulness of this assumption.
It would be interesting to see if, as it is reasonable to assume, audience prefers to select ideologically consistent news also because they rate this kind of content higher in quality and credibility. Further research should investigate the possibility that perception of news quality might be important in selecting the news, which can help the economical sustainability for the future of these media corporations.
To the best of my knowledge, there is no study on perceived news quality in the Italian media environment. The current research might contribute therefore to the beginning of cross national, comparative analysis on the European scale in order to shed light on practical implication for media economy and contribute to implementing news perception in communication studies on source cues.
This study has several limitations. In addition to the abovementioned ones, the sample of my study was quite small and not representative of the Italian population. Moreover, some additional concerns regarding representativeness have to be raised also regarding the methods of convenience and snowball sampling. These restricted the sampling in terms of age
31 (younger than the average of the country), geographical location (mostly from the Italian region of Tuscany) educational level and political knowledge. It is therefore important to state once again that cautiousness is required when drawing generalization beyond the study
participants. Another limitation lies in the fact that the article stimulus used for this
experiment dealt with the biography of a politician in office at the time of the data collection. Respondents might have answered in a biased way for some indicators (mostly in the liking category), according to the opinion that they already had of the President of the Republic. Moreover, the study was conducted as an online experiment. Therefore, participants
contributed to the study in their own environment and at their convenience, which might have brought some discrepancies to the data.
In conclusion, this research explored the effects that source cues have on news perception of online news content. The labels and the layout of a media corporations are not solely responsible for triggering a certain level of news quality, although they are still the most important cue to convey credibility in journalistic information. Readers will therefore still evaluate the credibility of a given news piece, even when it comes from an outlet which is ideologically distant from their own political perception. The importance of this study lies in the fact that it was the first one to use perception of news quality in trying to give a rationale for audience’s news selection. In sum, even if ideological distance did not affect news quality perception, this study showed that the audience do not solely evaluate content based on perceived ideological congruence.
32 References
Bennett, W., & Iyengar, S. (2008). A New Era of Minimal Effects? The Changing
Foundations of Political Communication. Journal Of Communication, 58(4), 707-731. doi:10.1111/j.1460-2466.2008.00410.x
Boukes, M., & Boomgaarden, H. (2014). Soft News With Hard Consequences? Introducing a Nuanced Measure of Soft Versus Hard News Exposure and Its Relationship With Political Cynicism. Communication Research, 42(5), 701-731. doi:10.1177/0093650214537520 Craig, D. (2011). Excellence in online journalism. Thousand Oaks, Calif.: SAGE.
De Vreese, C. H & Lecheler, S.K. (2012). News framing and public opinion: A mediation analysis of framing effects on political attitudes. Journalism & mass communication quarterly, 89 (2), 185-204. doi: doi:10.1177/1077699011430064
Druckman, J. (2001). On the Limits of Framing Effects: Who Can Frame?. Jop, 63(04). doi:10.1111/0022-3816.00100
Durante, R., & Knight, B. (2012). Partisan Control, Media Bias, and Viewer Responses: Evidence From Berlusconi’s Italy. Journal Of The European Economic
Association, 10(3), 451-481. doi:10.1111/j.1542-4774.2011.01060.x
Erikson, R., & Tedin, K. (2003). American Public Opinion - Its Origins, Content, and Impact (6th ed.). New York: Longman.
Esser, F., & Umbricht, A. (2013). Competing models of journalism? Political affairs coverage in US, British, German, Swiss, French and Italian newspapers. Journalism, 14(8), 989-1007. doi:10.1177/1464884913482551
Gentzkow, M., & Shapiro, J. (2006). Media Bias and Reputation. SSRN Journal. doi:10.2139/ssrn.642362
Gentzkow, M., & Shapiro, J. (2015). Ideology and online news. Cambridge, Mass.: National Bureau of Economic Research.
Go, E., Jung, E., & Wu, M. (2014). The effects of source cues on online news
perception. Computers In Human Behavior, 38, 358-367. doi:10.1016/j.chb.2014.05.044 Goren, P., Federico, C., & Kittilson, M. (2009). Source Cues, Partisan Identities, and Political
Value Expression. American Journal Of Political Science, 53(4), 805-820. doi:10.1111/j.1540-5907.2009.00402.x
Habermas, J. (1984). The theory of communicative action. Boston: Beacon Press. Hallin, D., & Mancini, P. (2004). Comparing media systems. Cambridge: Cambridge
University Press. Halperin, S., & Heath, O. (2012). Political research. Oxford: Oxford University Press.
Hibberd, M. (2007). Conflicts of interest and media pluralism in Italian broadcasting. West European Politics, 30(4), 881-902. doi:10.1080/01402380701500363
Hovland, C., & Weiss, W. (1951). The Influence of Source Credibility on Communication Effectiveness. Public Opinion Quarterly, 15(4), 635. doi:10.1086/266350
33 Iyengar, S., & Hahn, K. (2009). Red Media, Blue Media: Evidence of Ideological Selectivity
in Media Use. Journal Of Communication, 59(1), 19-39. doi:10.1111/j.1460-2466.2008.01402.x
Iyengar, S., Hahn, K., Krosnick, J., & Walker, J. (2008). Selective Exposure to Campaign Communication: The Role of Anticipated Agreement and Issue Public Membership. Jop, 70(01). doi:10.1017/s0022381607080139
Jerit, J., & Barabas, J. (2012). Partisan Perceptual Bias and the Information Environment. The Journal Of Politics, 74(03), 672-684. doi:10.1017/s0022381612000187
Kang H, Bae K, Zhang S and Sundar SS. (2011). Source cues in online news: Is the proximate source more powerful than distal sources? Journalism Mass Communication Q 2011; 88(04), 719–736.
Koh, Y., & Sundar, S. (2010). Effects of specialization in computers, web sites, and web agents on e-commerce trust. International Journal Of Human-Computer Studies, 68(12), 899-912. doi:10.1016/j.ijhcs.2010.08.002
Levendusky, M. (2010). The partisan sort. Chicago: University of Chicago Press. Levine, D., & Newman, N. (2014). 2014 Reuter Institute Digital News Report. Oxford:
Reuter Institute for the Study of Journalism. Retrieved from
http://reutersinstitute.politics.ox.ac.uk/sites/default/files/Reuters%20Institute%20Digital% 20News%20Report%202014.pdf
Lucassen, T., & Schraagen, J. (2011). Factual accuracy and trust in information: The role of expertise. Journal Of The American Society For Information Science And Technology, 62(7), 1232-1242. doi:10.1002/asi.21545
Lucassen, T., & Schraagen, J. (2012). Propensity to trust and the influence of source and medium cues in credibility evaluation. Journal Of Information Science, 38(6), 566-577. doi:10.1177/0165551512459921
Mancini, P. (2000). How to Combine Media Commercialization and Party Affiliation: The Italian Experience. Political Communication, 17(4), 319-324.
doi:10.1080/10584600050178906
Mancini, P. (2013). The Italian public sphere: a case of dramatized polarization. Journal Of Modern Italian Studies, 18(3), 335-347. doi:10.1080/1354571x.2013.780348
Massaro, D., Petty, R., & Cacioppo, J. (1988). Communication and Persuasion: Central and Peripheral Routes to Attitude Change. The American Journal Of Psychology, 101(1), 155. doi:10.2307/1422805
McQuail, D. (1992). Media performance. London: Sage Publications.
Messing, S., & Westwood, S. (2012). Selective Exposure in the Age of Social Media: Endorsements Trump Partisan Source Affiliation When Selecting News
Online. Communication Research, 41(8), 1042-1063. doi:10.1177/0093650212466406 Metzger, M., & Flanagin, A. (2013). Credibility and trust of information in online
environments: The use of cognitive heuristics. Journal Of Pragmatics, 59, 210-220. doi:10.1016/j.pragma.2013.07.012
34 Murphy, C., & Westbury, C. (2013). Expanding the Scope of Selective Exposure: An
Objective Approach to Measurement of Media Ideology. Communication Methods And Measures, 7(3-4), 224-246. doi:10.1080/19312458.2013.813921
O'Sullivan, J., & Heinonen, A. (2008). OLD VALUES, NEW MEDIA. Journalism Practice, 2(3), 357-371. doi:10.1080/17512780802281081
Plasser, F. (2005). From Hard to Soft News Standards?: How Political Journalists in Different Media Systems Evaluate the Shifting Quality of News. The Harvard International Journal Of Press/Politics, 10(2), 47-68. doi:10.1177/1081180x05277746
Ragnedda, M. (2014). Censorship and media ownership in Italy in the Era of Berlusconi. Global Media Journal: Mediterranean Edition, 9(1), Spring 2014, 13 – 26.
Rogers, E., & Kincaid, D. (1981). Communication networks. New York: Free Press.
Schuck, A., & Feinholdt, A. (2015). News Framing Effects and Emotions. Emerging Trends In The Social And Behavioral Sciences, 1-15. doi:10.1002/9781118900772.etrds0241 Sears, D. (1986). College sophomores in the laboratory: Influences of a narrow data base on
social psychology's view of human nature. Journal Of Personality And Social Psychology, 51(3), 515-530. doi:10.1037/0022-3514.51.3.515
Severin, W., & Tankard, J. (1979). Communication theories. New York: Hastings House. Snow, R., & Graber, D. (1986). Processing the News: How People Tame the Information
Tide. Contemporary Sociology, 15(1), 73. doi:10.2307/2070921
Stroud, N. (2007). Media Use and Political Predispositions: Revisiting the Concept of Selective Exposure. Polit Behav, 30(3), 341-366. doi:10.1007/s11109-007-9050-9 Stroud, N. (2010). Polarization and Partisan Selective Exposure. Journal Of
Communication, 60(3), 556-576. doi:10.1111/j.1460-2466.2010.01497.x
Sundar, S. (1999). Exploring Receivers' Criteria for Perception of Print and Online News. Journalism & Mass Communication Quarterly, 76(2), 373-386.
doi:10.1177/107769909907600213
Sundar, S. (2001). Conceptualizing sources in online news. Journal Of Communication, 51(1), 52-72. doi:10.1093/joc/51.1.52
Sundar, S. (2008) “The MAIN Model: A Heuristic Approach to Understanding Technology Effects on Credibility." Digital Media, Youth, and Credibility. Edited by Miriam J. Metzger and Andrew J. Flanagin. The John D. and Catherine T. MacArthur Foundation Series on Digital Media and Learning. Cambridge, MA: The MIT Press, 73–100. doi: 10.1162/dmal.9780262562324.073
Sundar, S., Knobloch-Westerwick, S., Hastall, M. (2007). News cues: Information scent and cognitive heuristics. Journal Of The American Society For Information Science And Technology,58(3), 366-378. doi:10.1002/asi.20511
Tworzecki, H., & Markowski, R. (2014). Knowledge and Partisan Bias: An Uneasy Relationship. East European Politics & Societies, 28(4), 836-862.
35 Valentino, N., Beckmann, M., & Buhr, T. (2001). A Spiral of Cynicism for Some: The
Contingent Effects of Campaign News Frames on Participation and Confidence in
Government. Political Communication, 18(4), 347-367. doi:10.1080/10584600152647083 Vraga, E., & Tully, M. (2015). Media Literacy Messages and Hostile Media Perceptions:
Processing of Nonpartisan Versus Partisan Political Information. Mass Communication And Society, 18(4), 422-448. doi:10.1080/15205436.2014.1001910
Williams, A. (2012). Trust or Bust?: Questioning the Relationship Between Media Trust and News Attention. Journal Of Broadcasting & Electronic Media, 56(1), 116-131.
doi:10.1080/08838151.2011.651186
36 Appendix
37 2. Il Giornale
38 3. La Repubblica
