The Influence of News Brand Cues and Story Content on Citizen Perceptions of News Bias

Abstract

Mainstream U.S. news media stand accused of bias against the forty-fifth president, Donald Trump. The relentlessness and intensity of these accusations over the course of Trump’s presidency are unusual and make for an opportunity to study perceptions of news bias. During the experiment reported here, participants (N = 315) were exposed to biased (pro- and anti-Trump) news stories that were attributed to either CNN, Breitbart, or remained unattributed to a news brand. After reading the stories, participants rated the stories for their relative slantedness in favor of, neutral, or against the president. Findings reveal that news users are sensitized to the president’s accusations of bias against CNN. For example, anti-Trump stories were rated as more slanted than pro-Trump stories when they were attributed to CNN. This was not the case when the same stories were attributed to Breitbart. Interestingly, unattributed biased news received the highest ratings for slantedness.

Keywords

experiment bias news slant ideological stance news brand political identity hostile media effect news selection polarization

The contemporary media landscape is often described in dystopian terms: information-saturated, polarized, and polluted with disinformation. In response, social and information scientists have studied the ways in which media users navigate this environment and identified a number of cognitive shortcuts that serve this process. Among these are heuristics that enable split-second decisions about the interest value of a piece of information and its credibility (Liu and Wei 2019; Winter et al. 2016). The origin or source from which information flows is a central cue in triggering heuristics for news selection. News outlets, most likely aware of heuristics, carefully brand the way they do journalism, offering subtle cues about their political leanings to attract niche audiences (Brems et al. 2017; Sjøvaag 2013). This dynamic has captured the attention of scholars who study news selection, showing that in line with the confirmation bias hypothesis, news users choose news that aligns with their political leanings, using content cues (Knobloch-Westerwick 2012), the reputation of the source (Iyengar and Hahn 2009), or endorsements on social media (Messing and Westwood 2014) to guide them. Yet, there is sparsity of insight into the processes that shape perceptions about news sources, which precede news selection behavior. Thus, as a prequel to news selection research (Kaye and Johnson 2016), our goal was to investigate how source and content cues interact in shaping news user perceptions of news bias.

The findings offer new insights on at least three levels. First, it reveals how news users weigh source attribution against actual bias in news stories to evaluate news objectivity. Second, the models we tested demonstrated the influence of individual-level heuristics in shaping perceptions of news bias. Third, to our knowledge, this is the first study to comparably assess the effects of branded and unbranded news on perceptions of news bias. Contemporary news use, facilitated by social media sharing, algorithmically derived newsfeeds, or bot activity, often happens with few or fake source cues (Kalogeropoulos et al. 2019; Lazer et al. 2018). Our study offers insight into how media users assess bias in unbranded news at a time when systematic evidence is paramount to facilitating the resilience of citizens against trolls and bots that are active in disseminating news on the social web.

News Brands as Heuristic Partisan Cues

Increased competition in media markets during an age of 24/7 cable news and soaring online news media options compelled news organizations to create distinct news brand identities, much like consumer products are marketed (e.g., Hamilton 2004). For example, CNN brands itself as the Most trusted name in News, whereas The Fox News Channel recently rebranded as Real news. Real honest opinion, replacing the twenty-year-old slogan Fair and balanced.

One successful branding strategy is to insert an ideological slant to attract a subset of the public who shares similar ideological views—thereby focusing on a market segment instead of appealing to a general public (Gentzkow et al. 2014). The Fox News Channel, for example, took this strategy to provide news that appeals to conservative viewers who saw mainstream media as biased against their political sentiments (Jones 2012; Kurtz 1999). As a consequence, ideological leanings of news platforms are integrated into brand identities, signaling to potential viewers what to expect from their reports (Gentzkow and Shapiro 2006).

This ideology-based branding strategy, however, takes its toll on how news users evaluate news objectivity in an increasingly polarized media environment. Studies on hostile media effects (HMEs) showed that partisans perceive news to be biased if it is attributed to a brand affiliated with a rival group (Ariyanto et al. 2007; Arpan and Raney 2003) or an opposing party (Hyun and Seo 2019; M. Kim 2019). Tully et al. (2020) conducted in-depth interviews with partisans after they read two news stories—one from Fox and the other from The New York Times—and reported that participants relied on the ideological reputation of news outlets to judge bias in stories. Evidence from experimental studies confirms that perceptions of news brand ideology influence how partisans evaluate the objectivity of news content. For example, self-identified partisans perceive the same news as more favorable of an opposing party when it is attributed to an outlet with a reputation of supporting that opposing party (Baum and Gussin 2008; Turner 2007). The findings of these studies advance the HME tenet that news brands offer a heuristic cue to infer the ideological slant of journalistic content. In the context of the current study, the same news content about President Trump may be perceived as more biased against him if it is attributed to CNN (known to be accused of anti-Trump sentiment) than Breitbart (widely regarded as pro-Trump). To test this prediction for a main effect for news brand, we formulated the following hypothesis:

Hypothesis 1 (H1): When news about Trump is attributed to CNN, it would be perceived as more biased against him than when the same content is attributed to Breitbart.

H1 rests on the idea that news brand names function as heuristic cues that influence the evaluation of news objectivity. What about unbranded news, void of a heuristic cue? Turner (2007) offers the only known experimental study that directly tested for the comparative impact of branded versus unbranded news on perceptions of bias. Turner (2007) used audiovisual news and reported that the unbranded condition was rated somewhere between CNN and The Fox News Channel in terms of bias. In line with this, there are indicators from the marketing literature that news users respond more favorably toward a news product, price, promotion, and distribution of branded news than generic or unbranded (fictitiously named or unnamed) products in the same category (Keller 2002). Yet, the question remains: How contemporary news users will evaluate the objectivity of textual, rather than audiovisual, news if it is unbranded? With no heuristic cue available, news users might be less distracted by the source cue and able to judge news content bias in an unfiltered way. Alternatively, in an overcrowded news landscape where alarms about fake news are frequently sounded, unbranded news might elicit a level of distrust in users that could sharpen suspicion about bias. Due to lack of concrete and recent evidence to steer a hypothesis, a research question was posed to guide analyses:

Research Question 1 (RQ1): Would news that is not attributed to a brand be perceived as less biased than news attributed to CNN or Breitbart?

From a democratic theory perspective, the automaticity of heuristic processing that is triggered by news brand attribution is seen as an incomplete account of how citizens assess news bias. Indeed, the view of citizens as rational participants in democratic life affords them the capacity to pick up the signals of ideological leaning in news content and form perceptions about news bias based on such evidence. Previous studies have offered support of this tenet. When a news article is ideologically slanted, partisans often reach agreement on the direction of bias but may disagree about the degree of slantedness (Gunther et al. 2001). These observations temper the view that news brand heuristic cues overwhelm news content cues in the process whereby citizens form perceptions about news bias (Turner 2007). Our study is well positioned to offer insight into the possibility that content bias would override the heuristic power of news brand cues during the process of scrutinizing news for bias. If content slantedness is a stronger driver in bias perception than news brand cues, pro-Trump and anti-Trump bias scores should be independent of news brand attribution. To test this, an interaction effect between the two experimental factors (message tone and news brand) will be examined. This interaction could also offer insights into the effects of incongruence between message tone and brand reputation.

In the case of our study, incongruence would emerge when pro-Trump news is attributed to an anti-Trump news brand, and vice versa. In the context of HME, only a handful of studies have examined the effects of message tone and source congruence on the assessment of news objectivity (e.g., Gunther et al. 2017; M. Kim 2019). The findings of these studies suggest that news users perceive (1) more bias when the expected slant of a news source is consistent with the tone of the news content and (2) less bias from the same news content when the expected ideological slant of the news source is at odds with the message tone. Similarly, a study on newspaper endorsements found that the official backing from a newspaper that is perceived to be leaning in the opposite ideological direction than a candidate increases voting intent for the candidate (Chiang and Knight 2011).

Valuable as these studies are in understanding the ramifications of (in)congruence between news brand and message tone, the interaction between these factors is in need of robust testing—specifically on perceptions of bias as an outcome. Both Gunther et al. (2017) and M. Kim (2019) employed study designs, which used a single issue and recruited partisans who held strong opinions on the story issues used as stimuli. Our work extends this line of inquiry by testing multiple stories among participants who varied in opinions about the topics featured in the messages. In doing so, we formulated the following hypothesis regarding the interaction between message tone and news brand on the assessment of news objectivity:

Hypothesis 2 (H2): The effects of news brand on perceived story slantedness will be moderated by message tone.

We expect anti-Trump messages will be viewed as more biased when they appear on CNN than Breitbart, whereas pro-Trump messages will be viewed as less biased when they are attributed to CNN than Breitbart. Overall, CNN’s perceived slantedness will probably be higher for anti- than pro-Trump messages—and Breitbart’s perceived slantedness will be higher for pro- than anti-Trump messages.

Political Party Identification and Views on News Objectivity

The accusation that media are biased against the Republican Party has persisted for decades (Lee 2005; Pew Research Center 2014). With the exception of Gerald Ford, Republican presidents since Richard Nixon have all charged the mainstream news media with liberal bias (Shafer 2003) which is generally not supported by research evidence (e.g., Grabe and Bucy 2009; Watts et al. 1999). Some scholars have argued that the strategy of accusing the media of bias has cultivated a decline in the public’s trust of American journalism (e.g., Smith 2010; van Duyn and Collier 2019). Polarized views on media objectivity continue to grow in the Trump era, with a steady stream of attacks on mainstream news outlets like CNN for producing fake news and acting as the enemy of the people. By contrast, the president and his spokespeople praise The Fox News Channel and Breitbart for balanced reporting (Boyle 2016; Shephard 2017).

Public discourse that constructs news outlets dichotomously as either liberal or conservative resonates with political party identities of news users and influences their assessments of bias. Indeed, the results of opinion polls about news media show that self-identified Republicans think almost all news media are biased, while non-Republicans see several news outlets as unbiased (Knight Foundation 2018). More than 60 percent of self-identified Republicans also reported cynical views on the role news media play in society, about a 40 percent increase from 2016 (Pew Research Center 2018). It is therefore reasonable to expect that Republicans would report more slantedness than self-identified Democrats and self-identified non-partisans. Based on this tendency, we formulated the following hypothesis:

Hypothesis 3 (H3): The effect of news brand attribution (CNN, Breitbart, unattributed) on perceived news slantedness will be stronger among self-identified Republicans than Democrats and non-partisans.

Individual Perceptions of News Brands

Based on the evidence of a polarized American news media environment, HME research generally assumes that the influence of news brands on user assessments of objectivity is similar across individual citizens (e.g., Coe et al. 2008; Gunther et al. 2017). These studies typically do not consider how fine-grain variations in individual perceptions of news brands guide political opinions and behavior (e.g., Kaye and Johnson 2016; Tsfati 2003). At the same time, political communication scholars have shown that preexisting cognitive and affective orientations toward a news brand influence the degree to which a person engages with news from a particular brand (Arendt et al. 2019; Stroud and Lee 2013). Prior attitudes and beliefs have also been shown to direct political information processing in a way that gives more weight to messages consistent with beliefs and attitudes (e.g., Taber and Lodge 2006).

In line with this, one could expect that an individual’s assessments of news about Trump published by Breitbart may be closely related to the extent that the news user sees Breitbart as pro-Trump or conservative. The same logic goes for the assessment of bias attributed to CNN, now cast as a liberal anti-Trump news brand. News attributed to CNN may be perceived as anti-Trump in line with a person’s perception of CNN as left-leaning which will be different for a person who sees CNN as a politically centrist news brand. Therefore, we expect that individual perceptions of where a news brand stands on the ideological continuum may color judgments about news objectivity. To test this idea, the following hypothesis was formulated:

Hypothesis 4 (H4): Individual perceptions about the ideological leanings of a news brand will drive perceptions of bias in slanted news stories.

Method

An online experiment with a 2 (message tone) × 3 (news brand) within-subjects design was conducted. The two levels of message tone were represented by pro- and anti-Trump messages. News brand was represented by CNN (anti-Trump), Breitbart (pro-Trump), and no news brand identity. Each experimental factor was represented twice to overcome the limitations associated with single message experiments. The experimental design and the procedures were approved by the Institutional Review Board of Indiana University (#1709357168).

Experimental Manipulation

Message tone

This within-subjects factor was represented by six pro- and six anti-Trump news stories, created from editorials published in professional news outlets, including CNN, The New York Times, The Washington Post, Fox News Channel, National Review, and Washington Examiner. Following previous studies that tested the effects of opinionated messages on assessment of news objectivity (e.g., Gunther et al. 2017), we created experimental stimuli using editorials as they tend to, compared with non-editorial news, take an ideological stance on issues. Stories were shortened to a maximum of two hundred words.¹ Because of the length reduction, stories lost the appearance of full-scale news editorials and resembled shorter stories that are typically seen on social media newsfeeds. To avoid sensitization to slantedness, participants were told that they will read excerpts from news stories. Online Appendix K presents all twelve stimuli used in the experiment.

News brand names

Each participant read four stories unattributed to a news brand, four attributed to Breitbart, and four attributed to CNN. Breitbart and CNN logos were placed in the upper-left corner of the masthead of a story to identify the two outlets that were selected based on their reputations for partisanship.² A survey conducted in March 2018 found 71 percent of Americans viewed CNN as biased, while 23 percent rated CNN as unbiased. By comparison, 68 percent of respondents saw Breitbart News as biased and 6 percent believed it was unbiased.³ Political leanings of respondents yielded stark differences in bias ratings. Among Republicans and Republican-leaners, CNN was considered the most biased news organization, with a −87 bias rating. By comparison, Democrats and Democratic-leaners gave CNN a +29—a 116 point partisan gap. Breitbart News received a −73 bias score among Democrats and Democrat-leaners and among Republicans and Republican-leaners. Breitbart earned a −19 rating—a 54 point spread among partisans (Knight Foundation 2018).

Measures

News brand ideological leanings

Participants reported their perceptions of nine news brands, including CNN and Breitbart, in terms of ideological leanings. Nine news brands were used to disguise the study’s objective of focusing on CNN and Breitbart while validating the news brand manipulation (see Online Appendix I for more details). Responses to these questions were measured on a 7-point semantic differential scale ranging from 1 (very liberal) to 7 (very conservative) and including a “Don’t know” option. About 15 percent of participants (n = 48) chose “Don’t know” in response to CNN. Breitbart evoked “Don’t know” responses in 38 percent (n = 120) of participants. A total of 40 participants (12.7%) chose the “Don’t know” option for both these brands. On average, participants perceived CNN (M = 2.69, SD = 1.49) as more liberal than Breitbart (M = 5.14, SD = 1.98), t(186) = −14.50, p < .01, lending support to the deployment of these two brands as ideologically polar in the minds of news users.⁴

Perceived story stance and slantedness

Three 7-point semantic differential items adopted from Arpan and Raney (2003) were administered after each story. The first item asked whether a story was extremely biased in favor (3) or against (−3) President Trump. The second item asked whether the story made President Trump seem extremely likable (3) or unlikable (−3). The third item asked whether the story made President Trump look extremely good (3) or bad (−3). Responses to these items were reliable (Cronbach’s α ≥ .84) as an index.

Perceived story stance

To measure the direction of perceived bias, individual responses to the three items were averaged for each story creating an index that ranged from −3 (biased against President Trump) to 3 (biased in favor of President Trump). Each participant had twelve perceived stance scores—one for each story (anti-Trump: M = −1.21, SD = 1.39; pro-Trump: M = 1.06, SD = 1.29).

Perceived story slantedness

The absolute value of each perceived stance score was used to create the story slantedness scale. The scale captured the intensity of perceived bias for each story, irrespective of direction. Values ranged from 0 (balanced) to 3 (slanted). Each individual had twelve perceived story slantedness scores (pro-Trump: M = 1.32, SD = 1.02; anti-Trump: M = 1.49, SD = 1.07).

Participants and Procedure

The participants in this study were recruited from a national paid opt-in online survey through Qualtrics Panels.⁵ The sample procedure included a quota to recruit an equal number of Democrats, Republicans, and non-partisans. A number of procedures were implemented to ensure data quality. Specifically, participating in this study on mobile devices was prohibited. Participants were excluded if they failed an attention check (which was administered right before the main experiment) or spent too little time (less than a third of the median completion time of first thirty participants) on the experiment (see Online Appendix C for more details).

Participants completed a pre-exposure questionnaire that contained questions about demography, party identification, and news brand ideological leanings. Exposure to twelve experimental stimuli followed. To prevent early sensitization to the study’s objective, the first four randomly assigned slanted news stories (two pro- and two anti-Trump) represented the unattributed condition. The following eight stories were also presented in random order and featured news brand logos. After reading each slanted news story, participants were prompted to answer three questions that measured their perception of story stance. Participants were debriefed about the objective of the study upon completion of the experiment. In total, 315 participants (females: n = 213, 67.6%; white: n = 273, 86.7%) completed the study. The average age of participants was 49.2 (SD = 17.70). See Online Appendix B for descriptive statistics.

Data Analysis

To test the hypotheses and answer the research question, repeated-measures analyses of variance (ANOVAs) were conducted. Individual responses to each story were collapsed based on the two experimental conditions (message tone and news brand). Individual demographic information was not included in the model because the repeated nature of this experimental design is expected to control for the effects of individual differences. Of note, the results of analyses using the general linear mixed model, which used individual observations instead of collapsed responses, yielded essentially the same results as reported here. Standard errors reported below were clustered at the individual participant level to account for within-individual correlations.

Results

H1

The forecast that a news brand conventionally seen as anti-Trump (CNN) will be perceived as more slanted against Trump than a news brand conventionally seen as pro-Trump (Breitbart) was tested through a repeated-measures ANOVA with perceived story stance as the dependent variable. As shown in Table 1, there was a significant main effect for news brand on perceived story stance, F(2, 1552) = 5.82, p < .005, η² = .008. On average, stories attributed to CNN (M = −0.17, SE = 0.05) were perceived to hold a stronger anti-Trump slant (marked by the negative mean) than stories attributed to Breitbart (M = 0.04, SE = 0.05; p < .001), supporting H1.⁶

Table 1.

Notable F Test Results on Perceived Story Stance and Perceived Story Slantedness.

Variables	df	F	p	η²
DV: Perceived story stance without PID
Message tone (A)	1, 1552	2,152.72	>.000	.586
News brand names (B)	2, 1552	5.8	.003	.008
A × B	2, 1552	16.54	>.000	.021
DV: Perceived story stance with PID
Message tone (A)	1, 312	738.63	>.000	.703
News brand names (B)	2, 621	8.74	>.000	.027
Party identification (C)	2, 312	1.46	.234	.009
A × B	2, 579	33.11	>.000	.103
A × C	2, 312	0.10	.906	.001
B × C	4, 621	0.62	.646	.003
A × B × C	4, 621	0.48	.751	.003
DV: Perceived story slantedness without PID
Message tone (A)	1, 1522	22.17	>.000	.014
News brand names (B)	2, 1522	19.50	>.000	.025
A × B	2, 1522	8.51	>.000	.011
DV: Perceived story slantedness with PID
Message tone (A)	1, 312	14.05	>.000	.043
News brand names (B)	2, 312	25.03	>.000	.075
Party identification (C)	2, 312	2.03	.133	013
A × B	2, 579	10.37	>.000	.035
A × C	2, 312	2.91	.056	.018
B × C	4, 621	1.70	.148	.011
A × B × C	4, 579	0.48	.748	.003

Note. DV = dependent variable; PID = party identification.

RQ1

RQ1 prompted a comparison of unbranded news to CNN and Breitbart on perceptions of bias. Two repeated-measures ANOVAs were conducted, one with perceived story stance and the other with perceived story slantedness as the dependent variable. For perceived story stance, a post hoc comparison on a significant main effect (reported for H1) showed that unbranded stories (M = −0.08, SE = 0.05) were not statistically different from CNN or Breitbart (see Table 1).

For perceived story slantedness, indicating the intensity of bias with values ranging from 0 (balanced) to 3 (slanted), there was a significant main effect for news brand, F(2, 1522) = 19.50, p < .001, η² = .025. A post hoc comparison with Bonferroni correction showed that unattributed news stories (M = 1.47, SE = 0.05) were rated as more slanted than stories attributed to either Breitbart (M = 1.28, SE = 0.05, p < .001) or CNN (M = 1.21, SE = 0.05, p < .001). These findings offer evidence of healthy skepticism about unidentified news sources among news users. On average, the difference between CNN and Breitbart’s perceived story slantedness was statistically non-significant. Yet, there were significant differences between the unattributed condition and Breitbart and CNN such that both news brands were rated as less slanted than the unattributed condition.

H2

H2 predicted that the effects of news brand on perceived story slantedness would be a function of message tone. To test this prediction, a repeated-measures ANOVA was conducted. There was a significant interaction between message tone and news brand on perceived story slantedness, F(2, 1522) = 8.51, p < .001, η² = .011. This interaction was illustrated in Figure 1. Post hoc analyses were performed to identify cells of significant variation. Comparisons with Bonferroni correction showed that, for anti-Trump messages, CNN (M = 1.37, SE = 0.06, p = .338) was not perceived to be more slanted than Breitbart (M = 1.28, SE = 0.06). By contrast, pro-Trump stories were perceived to be more slanted when attributed to Breitbart (M = 1.30, SE = 0.06) than CNN (M = 1.06, SE = 0.05, p < .001). Moreover, pro-Trump stories were perceived to be less biased than anti-Trump stories when attributed to CNN (p < .001). The difference between anti- and pro-Trump stories in the Breitbart condition was statistically insignificant (p = .721).⁷ These findings partially support H2, as the incongruence between message tone and expected ideological slant of news brand significantly dampened perceptions of story bias for pro-Trump messages attributed to CNN. This mix of findings does not provide conclusive evidence about the relative power of message tone compared with news brand cues in the process of forming perceptions of news bias. Pro-Trump messages varied in perceived bias levels depending on the news brand that they were attributed to.⁸

Figure 1.

Interaction between message tone and news brand names on perceived story slantedness.

H3

H3 required a test of self-reported party identification as a moderator for the impact of news brands on story slantedness. A repeated-measures ANOVA was performed using news tone, news brand names, and self-reported party identification as the independent variables. Contrary to expectation, party identification did not moderate the effects of either message tone or news brand on perceived story slantedness. Self-identified Republicans were not more sensitive to either message tone or news brand than self-identified Democrats or non-partisans (see Online Appendix D for the means and standard deviations of each condition).

H4

H4 predicted that perceived slantedness in news story content will be driven by preexisting individualized perceptions of how news brands are ideologically aligned. This prediction was tested comparing the results of two repeated-measures ANOVAs with and without measured news brand ideological leanings as the moderating variable. Because the perceived ideological leanings of the unbranded condition cannot be measured, the data analysis only focused on CNN and Breitbart stories. As shown in Table 2, there was a significant interaction effect between message tone and news brand factors, F(1, 916) = 16.26, p < .001, η² = .017, before including individual-level perceptions of ideological leanings for each news brand as a covariate in that model. From Table 2, it is clear that including individual-level perceptions of news brand ideological leanings into the model rendered the message tone by news brand interaction statistically non-significant. Moreover, while the main effects of perceived news brand ideological leanings were not significant, it produced a statistically significant interaction with message tone on perceived story slantedness, F(1, 914) = 18.46, p < .001, η² = .020.

Table 2.

Comparison between Perceived Ideological Leaning of News Brands on Story Slantedness.

Variables	df	F	p	η²
Perceived story slantedness without perceived ideological leaning
Message tone (A)	1, 916	12.70	>.000	.014
News brand names (B)	1, 916	2.94	.087	.003
A × B	1, 916	16.26	>.000	.017
Perceived story slantedness with perceived ideological leanings
Message tone (A)	1, 914	28.02	>.000	.030
News brand names (B)	1, 914	1.39	.238	.001
Perceived ideological leanings (C)	1, 914	0.01	.920	.000
A × B	1, 914	0.16	.686	.000
A × C	1, 914	18.46	>.000	.020

Note. Perceived ideological leanings = “Perceived news brand ideological leanings (1 = Very Liberal to 7 = Very Conservative).”

Close examination of the interaction showed that, for pro-Trump stories, the perceived news brand ideological leanings (1 = very liberal to 7 = very conservative) had a positive association with perceived story slantedness. Specifically, a standard deviation increase in perceived news brand ideological leaning (about 1.93) is associated with a .11 increase in perceived story slantedness (b = .11, p = .02). By contrast, for anti-Trump stories, a standard deviation increase in perceived news brand ideological leanings was associated with a decrease of about .12 in perceived story slantedness. These findings are consistent with the prediction of H2. News users perceived more bias when message tone and perceived ideological stance of news brands were congruent. Thus, the more conservative the individualized view of the news brand, the more slanted the pro-Trump content was perceived to be. The opposite was true when the tone and expected ideological stance of news brands were incongruent. Put differently, the more liberal the news brand rating, the less slanted participants thought pro-Trump content was. These findings imply that individual-level perceptions of ideological leanings of news brands account for the effect of the dichotomous news brand experimental factor (CNN vs. Breitbart) on perceived story slantedness. Figure 2 illustrates changes in perceived story slantedness as a function of perceived news brand ideological leanings. H4 was supported.

Figure 2.

Interaction between message tone and perceived news brand ideological leaning on perceived editorial slantedness.

Discussion

The central goal of this study was to assess whether perceptions of deliberately slanted news would vary depending on the news brand they are attributed to. In short, perceptions varied across news brands, offering evidence that news attribution triggers heuristics that shape perceptions of slantedness in content. The most direct support came from a simple main effect for news brand: A Breitbart logo made stories appear more favorable to Trump than a CNN logo. An interaction effect for news brand and message tone added more nuance to this finding. Pro-Trump messages were perceived as more biased when they were attributed to Breitbart than to CNN while CNN was seen as the most objective news brand when it featured pro-Trump stories. We also found that individual-level perceptions of news brand stance amplified the influence of incongruence between the ideological reputation of a news brand and the content it produces. Specifically, congruence increased perceived slantedness whereas incongruence lowered perceptions of story slantedness. Taken together, these findings lead us to conclude that news brand heuristics and message content cues work in concert to form perceptions of news slantedness.

These interaction effects also offer insights into the potential impact of Donald Trump’s publicized views of CNN and Breitbart. News users might be sensitized to the Trump accusation of CNN as hostile to his presidency. At the same time, citizens have not assimilated these accusations wholesale and they are not oblivious to the president’s ties to Breitbart News. Two of the findings mentioned above lead us to this conclusion. First, when CNN is critical of the president, participants in this study saw the cable news outlet as highly slanted. Yet, when CNN was associated with pro-Trump news, participants recognized the positive slant as inconsistent with accusations from the president and rated the reporting as least slanted of any experimental condition. Second—and related to this pattern—pro-Trump stories were rated as more slanted when associated with Breitbart than CNN.

In the absence of news brand cues, news users displayed a healthy level of skepticism. Unbranded news was perceived as more slanted than CNN and Breitbart. In an age when the alarm is often and loudly sounded on disinformation, fake news, and media hoaxes, this is not surprising. Unbranded news triggered responses in line with surveys that report high levels of concern among citizens about the sidedness of information they encounter on the social web. Only 4 percent of respondents in a Pew Study (Pew Research Center 2019a) expressed no level of concern about the slantedness of information on social media and in another study, fake news was ranked as the fifth largest problem in the United States—on par with socioeconomic inequality and violent crime (Pew Research Center 2019b). While skepticism is no stand-in for media literacy when it comes to detecting disinformation, these surveys reveal public awareness that likely reverberates in our findings that unbranded news is seen as slanted. No wonder that fake news sites mimic the logos and names of established professional news sources around the world (Ruddick 2017).

Although several public opinion surveys have consistently demonstrated that Republicans are more likely than Democrats or non-partisans to view news media as biased against their ideological positions, our findings do not support this tendency. Self-identified Republicans were no more sensitive to message tone or news brands than self-identified Democrats or non-partisans. Thus, the political leanings of participants were not significant in predicting perceptions of news slant. Instead, fine-grain individual-level perceptions of a news brand’s stance emerged as a predictor of news user perceptions of slantedness in news. We tested only two news brands, but the findings about individual-level perceptions are likely applicable to other news brands. On this point, our study contributes to the two largely independent bodies of HME and Media Bias research by linking cultivated perceptions of news brands (Media Bias focus) and perceptions of news content bias (HME focus). Our findings confirm the power of acquired heuristics in swaying the minds of citizens about news objectivity. At the same time, the (in)congruence between news brand reputation and its content adds nuanced qualification to perceptions of news bias. Our findings offer reason to pursue future HME investigations on how perceptions of news bias influence perceptions of news brands.

The current study was conducted in the United States, a highly polarized country in terms of media use (Boxell et al. 2020). Moreover, the stimuli centrally featured the U.S. president, a leader who is relentless in expressing polarizing views about news outlets—thereby deepening the ideological divide that has developed over a number of decades (Bump 2018). The dynamics of a polarized society with a president who regularly attacks specific media brands are lamented for its impact on democratic life. But it provides a unique opportunity to conduct experimental research on the heuristics that drive polarization in media use. These sociopolitical dynamics also limit the applicability of our findings to other democratic systems. It is unlikely, for example, that some of our results would be replicated in a Scandinavian country like Norway, with a political leaning score (indicating the partisanship of the country’s online news audience) less than half the size of the U.S. score (Fletcher and Jenkins 2019). At the same time, our approach to include individual-level perceptions of news brand stance would go a long way in controlling for cross-national differences in political dynamics. Future cross-national comparative work would be beneficial to understand the mechanics of news brand heuristics in less polarized societies. Moreover, work on this trajectory might uncover unconsidered dimensions of news brand heuristics using stimuli about less controversial and polarizing political figures. In particular, a study that examines how the political affiliation of unknown politicians might interact with news brands and news story leanings would be a terrific next step to what we reported here.

The fact that a sizable number of participants were unaware of the ideological stance of Breitbart is potentially problematic, despite the evidence that low reported awareness of Breitbart’s political leanings had little impact on the findings (see Online Appendix F). News brands with well-known ideological leanings such as Fox News might have even stronger effects on perceived news bias. Thus, future work using an ideological range of news brands and featuring an assortment of news topics could add nuance to knowledge about how news brands function as heuristic cues in evaluations of news objectivity.

Our experimental procedure was an unlikely match for how media users typically engage with news. For example, none of the slanted news stories was attributed to identifiable authors, but for good reason. The ideological stance of news writers has been shown to influence perceived news bias (e.g., Gunther et al. 2017) and would therefore have introduced a source of unintended variance into our design. Also, forced choice experiments have been associated with overestimations of news bias (Arceneaux et al. 2012). A more natural news consumption scenario would increase the ecological validity of findings but proves to be a challenge for experimental control.

Our findings are not applicable to audiovisual, audio, or static image plus text news. Modality of content (textual vs. audiovisual) has been shown to influence how news users process information (Grabe et al. 2009). Modalities may also vary in how they afford news brand names to function as heuristic cues. Given the dearth of cross-modal research, we advocate for future work that looks comparatively at how news modalities elicit heuristic cues in forming perceptions of objectivity. Finally, the current study introduced an individual-level measure of news brand reputation into analyses. Yet, perceptions about the ideological reputation of a news brand most likely involve multi-dimensional complexity (J. Kim et al. 2010). More work is needed. Studies that test the inter-relationship between message dimensions (e.g., credibility, accuracy, trustworthiness; see Stroud and Lee 2013) and individual-level variables (e.g., news use, personal traits) would advance a more comprehensive account of what is in play when a news brand heuristic is activated in an individual citizen’s assessment of news objectivity.

Given the news media’s responsibilities as the fourth estate in democratic governance, diminishing trust in their evenhandedness to inform the public is seen as a core indicator of the democratic crises that a choir of public and academic authors write and speak about. The experiment reported here offers a stepping stone in understanding the factors involved in shaping perceptions of news slant. At a time when populist world leaders routinely use attacks on news media as a means of camouflaging their failures and polarizing their nations, the ability of citizens to distinguish slanted from non-slanted news becomes increasingly important. The study reported here provides some indications that citizens are able to recognize news reportage as slanted, but heuristic shortcuts about news platforms cloud that acuity.

Supplemental Material

RA-IJPP-Nov-2019-199.R2_Final_Supplementary_Information__online_supp – Supplemental material for The Influence of News Brand Cues and Story Content on Citizen Perceptions of News Bias

Supplemental material, RA-IJPP-Nov-2019-199.R2_Final_Supplementary_Information__online_supp for The Influence of News Brand Cues and Story Content on Citizen Perceptions of News Bias by Minchul Kim and Maria Elizabeth Grabe in The International Journal of Press/Politics

Footnotes

Declaration of Conflicting Interests

The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The author(s) received no financial support for the research, authorship, and/or publication of this article.

ORCID iD

Minchul Kim

Supplemental Material

Supplemental material for this article is available online.

Notes

Author Biographies

Minchul Kim (PhD, Indiana University) is a Research Fellow in the Institute for Convergence at Sungkyunkwan University. He is interested in political communication and public opinion. His research investigates how digital media platforms influence political communication processes.

Maria Elizabeth Grabe (PhD, Temple University) is a Member of the Observatory on Social Media and a Professor in the Media School at Indiana University. She does research at the intersection of news user demography (social class and gender) and multimodal message variables to understand informed citizenship.

References

Arceneaux

Kevin

Johnson

Martin

Murphy

Chad

. 2012. “Polarized Political Communication, Oppositional Media Hostility, and Selective Exposure.” The Journal of Politics 74 (1): 174–86.10.1017/S002238161100123X

Arendt

Florian

Northup

Temple

Camaj

Lindita

. 2019. “Selective Exposure and News Media Brands: Implicit and Explicit Attitudes as Predictors of News Choice.” Media Psychology 22 (3): 526–43.10.1080/15213269.2017.1338963

Ariyanto

Amarina

Hornsey

Matthew J.

Gallois

Cindy

. 2007. “Group Allegiances and Perceptions of Media Bias: Taking into Account Both the Perceiver and the Source.” Group Processes & Intergroup Relations 10 (2): 266–79.10.1177/1368430207074733

Arpan

Laura M.

Raney

Arthur A.

2003. “An Experimental Investigation of News Source and the Hostile Media Effect.” Journalism & Mass Communication Quarterly 80 (2): 265–81.10.1177/107769900308000203

Baum

Matthew A.

Gussin

Phil

. 2008. “In the Eye of the Beholder: How Information Shortcuts Shape Individual Perceptions of Bias in the Media.” Quarterly Journal of Political Science 3 (1): 1–31.10.1561/100.00007010

Boxell

Levi

Gentzkow

Matthew

Shapiro

Jesse M.

2020. Cross-Country Trends in Affective Polarization” the National Bureau of Economic Research. https://www.nber.org/papers/w26669.

Boyle

Matthew

. 2016. “President-Elect Donald J. Trump Praises Breitbart News Network: ‘It’s a Pretty Big Thing.’” Breitbart, November 22. https://www.breitbart.com/politics/2016/11/22/president-elect-donald-j-trump-praises-breitbart-news-network-its-a-pretty-big-thing/.

Brems

Cara

Temmerman

Martina

Graham

Todd

Broersma

Marcel

. 2017. “Personal Branding on Twitter: How Employed and Freelance Journalists Stage Themselves on Social Media.” Digital Journalism 5 (4): 443–59.10.1080/21670811.2016.1176534

Bump

Philip

. 2018. “Trump Makes It Explicit: Negative Coverage of Him Is Fake Coverage.” The Washington Post, May 5. https://www.washingtonpost.com/news/politics/wp/2018/05/09/trump-makes-it-explicit-negative-coverage-of-him-is-fake-coverage/?utm_term=.1f7a2d1ca8ed.

10.

Chiang

C. F.

Knight

2011. “Media Bias and Influence: Evidence from Newspaper Endorsements.” The Review of Economic Studies 78 (3): 795–820.10.1093/restud/rdq037

11.

Coe

Kevin

Tewksbury

David

Bond

Bradley J.

Drogos

Kristin L.

Porter

Robert W.

Yahn

Ashley

Zhang

Yuanyuan

. 2008. “Hostile News: Partisan Use and Perceptions of Cable News Programming.” Journal of Communication 58 (2): 201–19.10.1111/j.1460-2466.2008.00381.x

12.

Fletcher

Richard

Jenkins

Joy

. 2019. “Polarisation and the News Media in Europe.” Reuters Institute. https://reutersinstitute.politics.ox.ac.uk/our-research/polarisation-and-news-media-europe.

13.

Gentzkow

Matthew

Shapiro

Jesse M.

2006. “Media Bias and Reputation.” Journal of Political Economy 114 (2): 280–316.10.1086/499414

14.

Gentzkow

Matthew

Shapiro

Jesse M.

Sinkinson

Michael

. 2014. “Competition and Ideological Diversity: Historical Evidence from US Newspapers.” American Economic Review 104 (10): 3073–3114.10.1257/aer.104.10.3073

15.

Grabe

Maria E.

Bucy

Eric P.

2009. Image Bite Politics: News and the Visual Framing of Elections. New York: Oxford University Press.

16.

Grabe

Maria E.

Kamhawi

Rasha

Yegiyan

Narine

. 2009. “Informing Citizens: How People with Different Levels of Education Process Television, Newspaper, and Web News.” Journal of Broadcasting & Electronic Media 53 (1): 90–111.10.1080/08838150802643860

17.

Gunther

Albert C.

Christen

Cindy T.

Liebhart

Janice L.

Chia

Stella C.-H.

2001. “Congenial Public, Contrary Press, and Biased Estimates of the Climate of Opinion.” Public Opinion Quarterly 65 (3): 295–320.10.1086/322846

18.

Gunther

Albert C.

McLaughlin

Bryan

Gotlieb

Melissa R.

Wise

David

. 2017. “Who Says What to Whom: Content versus Source in the Hostile Media Effect.” International Journal of Public Opinion Research 29 (3): 363–83.10.1093/ijpor/8.3.248

19.

Hamilton

James T.

2004. All News that Fits to Sell. Princeton: Princeton University Press.

20.

Hyun

Ki D.

Seo

Mihye

. 2019. “The Effects of HMP and TPP on Political Participation in the Partisan Media Context.” Communication Research. Published electronically January 8. doi:10.1177/0093650218820229.

21.

Iyengar

Shanto

Hahn

Kyu S.

2009. “Red media, Blue Media: Evidence of Ideological Selectivity in Media Use.” Journal of Communication 59 (1): 19–39.10.1111/j.1460-2466.2008.01402.x

22.

Jones

Jeffrey P.

2012. “Fox News and the Performance of Ideology.” Cinema Journal 51 (4): 178–85.10.1353/cj.2012.0073

23.

Kalogeropoulos

Antonis

Fletcher

Richard

Nielsen

Rasmus K.

2019. “News Brand Attribution in Distributed Environments: Do People Know Where They Get Their News?” New Media & Society 21 (3): 583–601.10.1177/1461444818801313

24.

Kaye

Barbara K.

Johnson

Thomas J.

2016. “Across the Great Divide: How Partisanship and Perceptions of Media Bias Influence Changes in Time Spent with Media.” Journal of Broadcasting & Electronic Media 60 (4): 604–23.10.1080/08838151.2016.1234477

25.

Keller

Kevin L.

2002. Strategic Brand Management: Building, Measuring, and Managing Brand Equity. Upper Saddle River: Prentice Hall.

26.

Kim

Jooyoung

Baek

Tae H.

Martin

Hugh J.

2010. “Dimensions of News Media Brand Personality.” Journalism & Mass Communication Quarterly 87 (1): 117–34.

27.

Kim

Mihee

. 2019. “The Role of Partisan Sources and Audiences’ Involvement in The Persuasive Press Inference.” International Journal of Public Opinion Research 31 (2): 382–92.

28.

Knight Foundation. 2018. “Perceived Accuracy and Bias in the News Media.” Knight Foundation, June 20. https://knightfoundation.org/reports/perceived-accuracy-and-bias-in-the-news-media.

29.

Knobloch-Westerwick

Silvia

. 2012. “Selective Exposure and Reinforcement of Attitudes and Partisanship before a Presidential Election.” Journal of Communication 62 (4): 628–42.10.1111/j.1460-2466.2012.01651.x

30.

Kurtz

Howard

. 1999. “Crazy Like a Fox: Question His News Judgment, Perhaps, but Never Underestimate Roger Ailes.” The Washington Post, March 26. https://www.washingtonpost.com/archive/lifestyle/1999/03/26/crazy-like-a-fox-question-his-news-judgment-perhaps-but-never-underestimate-roger-ailes/a959339f-acaf-4823-8238-27da13eb77be/?utm_term=.de1544f1ef25.

31.

Lazer

David M. J.

Baum

Matthew A.

Benkler

Yochai

Berinsky

Adam J.

Greenhill

Kelly M.

Menczer

Filippo

Metzger

Miriam J.

Nyhan

Brendan

Pennycook

Gordon

Rothschild

David

Schudson

Michael

Sloman

Steven A.

Sunstein

Cass R.

Thorson

Emily A.

Watts

Duncan J.

Zittrain

Jonathan L.

2018. “The Science of Fake News.” Science 359 (6380): 1094–96.10.1126/science.aao2998

32.

Lee

Tien-Tsung

. 2005. “The Liberal Media Myth Revisited: An Examination of Factors Influencing Perceptions of Media Bias.” Journal of Broadcasting & Electronic Media 49 (1): 43–64.10.1207/s15506878jobem4901_4

33.

Liu

Bingjie

Wei

Lewen

. 2019. “Machine Authorship In Situ: Effect of News Organization and News Genre on News Credibility.” Digital Journalism 7 (5): 635–57.10.1080/21670811.2018.1510740

34.

Messing

Solomon

Westwood

Sean J.

2014. “Selective Exposure in the Age of Social Media: Endorsements Trump Partisan Source Affiliation When Selecting News Online.” Communication Research 41 (8): 1042–63.10.1177/0093650212466406

35.

Pew Research Center. 2014. “Political Polarization and Media Habits.” Pew Research Center, October 21. http://pewrsr.ch/1vZ9MnM.

36.

Pew Research Center. 2018. “Partisans Remain Sharply Divided in Their Attitudes about the News Media.” Pew Research Center, September 25. https://pewrsr.ch/2pqyCOQ.

37.

Pew Research Center. 2019a. “Americans Are Wary of the Role Social Media Sites Play in Delivering the News.” Pew Research Center, October 2. https://pewrsr.ch/2ZVuECq.

38.

Pew Research Center. 2019b. “Many Americans Say Made-Up News Is a Critical Problem that Needs to Be Fixed.” Pew Research Center, June 5. https://pewrsr.ch/2Xgk4At.

39.

Ruddick

Graham

. 2017. “Experts Sound Alarm over News Websites’ Fake News Twins.” The Guardian, August 18. https://www.theguardian.com/technology/2017/aug/18/experts-sound-alarm-over-news-websites-fake-news-twins.

40.

Shafer

Jack

. 2003. “The Varieties of Media Bias, Part 1: Who Threw the First Punch in the Press Bias Brawl?” Slate, February 3. https://slate.com/news-and-politics/2003/02/the-press-bias-brawl.html

41.

Shephard

Alex

. 2017. “Donald Trump Is Treating Fox News Like It’s State TV.” The New Republic. https://newrepublic.com/minutes/143564/donald-trump-treating-fox-news-like-its-state-tv.

42.

Sjøvaag

Helle

. 2013. “Revenue and Branding Strategy in the Norwegian News Market.” Nordicom Review 33 (1):53–66.10.2478/nor-2013-0004

43.

Smith

Glen R.

2010. “Politicians and the News Media: How Elite Attacks Influence Perceptions of Media Bias.” The International Journal of Press/Politics 15 (3): 319–43.

44.

Stroud

Natalie J.

Lee

Jae K.

2013. “Perceptions of Cable News Credibility.” Mass Communication and Society 16 (1): 67–88.10.1080/15205436.2011.646449

45.

Taber

Charles S.

Lodge

Milton

. 2006. “Motivated Skepticism in the Evaluation of Political Beliefs.” American Journal of Political Science 50 (3): 755–69.10.1111/j.1540-5907.2006.00214.x

46.

Tsfati

Yariv

. 2003. “Media Skepticism and Climate of Opinion Perception.” International Journal of Public Opinion Research 15 (1): 65–82.10.1093/ijpor/15.1.65

47.

Tully

Melissa.

Vraga

Emily K.

Smithson

Anne-Bennett

. 2020. “News Media Literacy, Perceptions of Bias, and Interpretation of News.” Journalism 21:209–26.10.1177/1464884918805262

48.

Turner

Joel

. 2007. “The Messenger Overwhelming the Message: Ideological Cues and Perceptions of Bias in Television News.” Political Behavior 29 (4): 441–64.10.1007/s11109-007-9031-z

49.

van Duyn

Emily

Collier

Jessica

. 2019. “Priming and Fake News: The Effects of Elite Discourse on Evaluations of News Media.” Mass Communication and Society 22 (1): 29–48.10.1080/15205436.2018.1511807

50.

Watts

Mark D.

Domke

David

Shah

Dhavan V.

Fan

David P.

1999. “Elite Cues and Media Bias in Presidential Campaigns: Explaining Public Perceptions of a Liberal Press.” Communication Research 26 (2): 144–75.10.1177/009365099026002003

51.

Winter

Stephan

Metzger

Metzger J.

Flanagin

Andrew J.

2016. “Selective Use of News Cues: A Multiple-Motive Perspective on Information Selection in Social Media Environments.” Journal of Communication 66 (4): 669–93.10.1111/jcom.12241

Supplementary Material

Please find the following supplemental material available below.

For Open Access articles published under a Creative Commons License, all supplemental material carries the same license as the article it is associated with.

For non-Open Access articles published, all supplemental material carries a non-exclusive license, and permission requests for re-use of supplemental material or any part of supplemental material shall be sent directly to the copyright owner as specified in the copyright notice associated with the article.

0.00 MB

0.71 MB