The inconsistency that is brought up in the previous paragraph leads to a conclusion that an independent analysis should be carried out in this paper. The analysis suggests a triangulation involving qualitative and quantitative content analysis, andtheory testing. This will maximize trustworthiness of this research as well asconfirm or reject its main hypothesis.

3.1 Quantitative content analysis: Sentimental vocabulary in media

A research conducted by Young and Soroka suggests that “a number of studies in political communication focus on the “sentiment” or “tone” of news contentSoroka S.N., Young L. Affective News: The Automated Coding of Sentiment in Political Texts [Electronic resource] // ResearchGate, 2012. URL: https://www.researchgate.net/publication/239798103_Affective_News_The_Automated_Coding_of_Sentiment_in_Political_Texts, P. 205 (Date of access: 15.05.2019).”. This paper is meant to be one of them as qualitative content analysis, based on word count, is an effective way to confirm or discard a hypothesis. Combined with other methods, it will provide a robust and trustworthy research paper. A dictionary-based Ibid. approach will be used in this paper. The dictionary consists of over 28 thousand negative and positive sentiments that will be automatically applied to two hundred articles in selected media outlets using Wordstat. This allows to find out whether the set of articles retrieved from specific media outlets use biased and emotive vocabulary that points to the role (side/stagefor debate) of each medium typeson the American political stage. A similar approach was used by D.T. Lowry Lowry D.T. Network Tv News Framing of Good Vs. Bad Economic News under Democrat and Republican Presidents: A Lexical Analysis of Political Bias // Journalism & Mass Communication Quarterly. 2008. V. 85. №3. P. 483-498. who used computerized quantitative content analysis to study “political news bias on network TV news Ibid. P.487.” and, among all, asserted that “a tiny but consistent pro-Democratic (and hence assumedly liberal) statement bias Ibid. P.484.”. However, this work is a bit outdated as it is now more than ten years old. Research by the Economist United States of Amoeba [Electronic resource] // The Economist, 2013. URL: https://www.economist.com/united-states/2013/12/05/united-states-of-amoeba (Date of access: 02.02.2019).shows that the main changes took place in 2013, when American congress turned into a polarized institution, while Trump's presidency has worsened the situation further with him becoming “the most polarizing president on record Bump P. Trump is the most polarizing president on record -- and almost nobody's opinion of him is changing [Electronic resource] // The Washington Post, 2019. URL: https://www.washingtonpost.com/politics/2019/01/16/trump-is-most-polarizing-president-record-almost-nobodys-opinion-him-is-changing/?noredirect=on&utm_term=.d8205c729e87 (Date of access: 15.05.2019).”: polarization has not only gained power in the U.S. Congress, but went beyond it and reached TV channels, newspapers, radio stations and the Internet. Consequently, abovementioned work requires an update, and research conducted in this paper is among all aimed at doing it.

3.1.1 Computerized Qualitative Content Analysis

QDA Miner package, precisely WordStat embedded into the software, will be used in the process of QCA. The tool allows to count words used in large arrays of text and visualizes statistics in the form of a table/bar chart/pie chart providing a vivid picture that allows researchers to analyze articles or interviews without counting words on their own. Word count is a crucial thing, and according to Hart and Childers, because “(a) word choices are important factor of news reporting…, (b) word choices aggregate into patterns of meaning…, and (c) patters of words accumulate in news consumers' minds… Hart R.P., Childers J.P. The Evolution of Candidate Bush // American Behavioral Scientist. 2005. V. 49. №2. P. 180-197.”.One of the main advantages of this approach is the fact that it is reliable in “the sense that they produce exactly the same results at the article level, whether one analyzes 10 articles or 10 years of newsSoroka S.N., Young L. Affective News: The Automated Coding of Sentiment in Political Texts [Electronic resource] // ResearchGate, 2012. URL: https://www.researchgate.net/publication/239798103_Affective_News_The_Automated_Coding_of_Sentiment_in_Political_Texts, P. 208 (Date of access: 15.05.2019).”.

In order to analyze articles by counting positive and negative sentiments a special dictionary must be used. In case of this specific work, the WordStat Sentiment Dictionary was used to accomplish this task. Main reasons to choose it are: a) it was updated just a year ago so its contents are up to date and do not require a revision, b) it combines more than 9,5 and 4,6 thousand negative and positive patterns respectively Sentiment Dictionaries [Electronic resource] // Provalis Research Inc., 2018. URL: https://provalisresearch.com/products/content-analysis-software/wordstat-dictionary/sentiment-dictionaries/ (Date of access: 15.05.2019).. This provides a good coverage of all connotations that may be used in selected articles. Finally, its modified version was used in one of the researches mentioned above, namely, the article by L. Young and S. SorokaSoroka S.N., Young L. Affective News: The Automated Coding of Sentiment in Political Texts [Electronic resource] // ResearchGate, 2012. URL: https://www.researchgate.net/publication/239798103_Affective_News_The_Automated_Coding_of_Sentiment_in_Political_Texts (Date of access: 15.05.2019).. The process will also be replicated using Hu and Liu's dictionary Opinion Lexicon Dictionary [Rar archive] // The University of Illinois at Chicago, 2004. URL: http://www.cs.uic.edu/~liub/FBS/opinion-lexicon-English.rar (Date of access: 15.05.2019). in order to enhance credibility of this part of triangulation.

3.1.2 CQCA: Data processing and results

WordStat in this case is used to count the number of words from a dictionary in uploaded texts, so uploading both dictionaries was the first thing to do. After that, one hundreddigital media articles were uploaded and checked. Results show that the most frequent negative word is “shutdown”. However, “shutdown” is in this specific case used as a topic for observation, not as a negative connotation. Subsequently, this negative sentiment must be excluded from the list of words.

First findings show that the frequency of negative sentiments across the board of digital media articles exceeded positive by ~47,5% (Figure 1,2).However, the real point of interest appears to be in “negative sentiments' rate per ten thousand words” which indicates that negative sentiments in Breitbart articles about government shutdown can be encountered at a rate that is lower than in HuffPost articles by a ~110% margin (Fig.3) and almost twice as low as in Politico articles (Ibid). In addition to that, WordStat has a co-occurrence tool and allows to search for coincidences of word use within one sentence or paragraph across all texts and is aimed at showing patterns of word use. The tool detected a high level of co-occurrences ofa negative word pattern (Fig. 4),that is used to describe Mexicanimmigrants to the U.S., plus the highest rate of interconnection between words in it. A significant share (around 70%) of such occurrences belongs to Breitbart articles (Fig. 5).

Figure 1. Positive and negative sentiments frequency in new media

Figure 2. Distribution of positive/negative sentiments in new media (frequency)

Figure 3. Negative sentiments in new media. Rate per 10'000 words

Figure 4. Co-occurrences in negative word patterns

Figure 5. Number of “Alien, Illegal, Criminal, Meant, Victim” pattern encounters by outlet

Opposite results can be observed in “positive sentiments' rate per ten thousand words” which shows that Huffington Post and Politico have a much lower rate of emotionally positive wording than Breitbart that seems to pay more attention to minimization of negative effects caused by government shutdown among its readers. This coincides with Breitbart's position on the ideological spectrum. The same can be said about Huffington Post and Politico, but more on that in the following subchapter that utilizes qualitative content analysis to check the hypothesis. Returning back to positive sentiments, Breitbart uses them45% and ~88% more frequently than Politico and HuffPost respectively (Fig. 6).

Figure 6. Positive sentiments in new media. Rate per 10'000 words

This seems to be the result of digital outlets' extreme level of polarization that was previously confirmed by research carried out on the Internet. When it comes to Breitbart, it was claimed to be far right on both resources, but HuffPost and Politico were in some cases rated as far left and left respectively, and exercised a left/slightly left bias in the other. Research conducted in this subchapter shows that reality is closer to the first scenario. In the case of government shutdown, the higher usage of negative sentiments plus lower usage of positive indicates that the outlet in question exercises left bias. As a result, Huffington Post, according to this research, exercises extremely left bias, Politico, while surely being much closer in both terms to Politico, can be thus rated as a “left”, and finally Breitbart is leaning far to the right, according to patterns found in this part of the paper.

Traditional media on the other hand have shown quite different results. Firstly, when it comes to distribution of sentiments old outlets demonstrate the approximately the same pattern as digital mediums: the difference between negative/positive sentiments usage is around 35% (Fig. 7)which is noticeably but not extremely different from what was seen in case of new media. What is extremely different from data seen before is the fact that in combination with lower difference between negative and positive sentiments' usage, distribution of connotations between media outlets does not replicate the pattern seen above (Fig. 8). Fox news, that is claimed to be partisan, shows a rate of negative sentiments that are quite similar to those seen on other media outlets, with only ABC (that is considered to exercise a left bias) falling behind the bulk by a small maximum margin of approximately16% (with average being 13,5%). Positive sentiments usageby old media (Fig. 9) haven't shown any interrelatedness between supposed ideological positions of any of the outlets and factual use of those sentiments. Fox is being almost on par with ABC (which has a left bias) while CBS and NBC (both exercise a slightly left bias) have the highest and the lowest usage of positive sentiments respectively. Co-occurrence tool has not found any significant ties between any positive or negative sentiments.

Figure 7. Distribution of sentiments in traditional media (Frequency)

Figure 8. Negative sentiments in old media. Rate per 10'000 words

Figure 9. Positive sentiments in old media. Rate per 10'000 words

Findings listed above indicate that distribution of negative and positive sentiments among old and new media are very different. On the one hand, new media section has shown that Internet mediums' political ideology/bias clearly correlates with the rate of positive/negative sentiments: while Breitbart, which is “far right”, abstains from using negatives and tends to place an emphasis on positive vocabulary in relation to the selected topic, Politico and Huffington Post, being “left” and “far left” show a completely opposite approach that favors negative connotations over positive. On the other hand, old media have not shown a clear link between their political ideology and the use of sentiments. Moreover, presumably one of the “extreme” media outlets, Fox News, shows a higher number of negative sentiments usage than its most liberal counterpart - ABC News.

In addition to that, the random sets of articles from both new and old media have shown that new mediums pay lots of attention to people who have little to do with politics. At the same time, those personalities can be quoted by opposing journalists, but in a negative context. For example, current random set of articles retrieved from each outlet using keywords “government shutdown”, has shown that Huffington Post uses rappers Richards K. Cardi B Has Some Advice For Trump Supporters During The Government Shutdown [Electronic resource] // Huffington Post, 2019. URL: https://www.huffpost.com/entry/cardi-b-government-shutdown-trump-supporters_n_5c439883e4b0a8dbe171f19a (Date of access: 15.05.2019). and comedy showsBaranauckas C. End Of Government Shutdown Is Horribly Timed For Comedy Shows [Electronic resource] // Huffington Post, 2019. URL: https://www.huffpost.com/entry/government-shutdown-comedy-shows_n_5c4bebebe4b06ba6d3bd1452 (Date of access: 15.05.2019). as news hooks for their published materials. At the same time, they completely ignore the other side of the debate. The opposing side actually refers to the same people, butuse negative subcurrents to deface their rivals'statements Hudson J. Hollywood Freaks Over Trump's Plan to Declare National Emergency: `A Giant Step Toward Dictatorship' [Electronic resource] // Breitbart, 2019. URL: https://www.breitbart.com/entertainment/2019/02/14/hollywood-freaks-over-trumps-plan-to-declare-national-emergency-a-giant-step-toward-dictatorship/ (Date of access: 15.05.2019); Caruso J. Cardi B Invited to Iowa Democart Event After Calling Trump Supporters `F*cking Racist Rednecks' [Electronic resource] // Breitbart, 2019. URL: https://www.breitbart.com/entertainment/2019/01/30/cardi-b-invited-to-iowa-democrat-event-after-calling-trump-supporters-fcking-racist-rednecks/ (Date of access: 15.05.2019); Caruso J. Watch: Rapper Cardi B Targets Trump in Endorsement of Bernie Sanders for President [Electronic resource] // Breitbart, 2019. URL: https://www.breitbart.com/entertainment/2019/04/15/watch-rapper-cardi-b-targets-trump-in-endorsement-of-bernie-sanders-for-president/ (Date of access: 15.05.2019); Caruso J. Rapper Offset Melts Over Government Shutdown: `That's Some Slave Sh*t' [Electronic resource] // Breitbart, 2019. URL: https://www.breitbart.com/entertainment/2019/01/23/offset-blasts-government-shutdown-thats-some-slave-sht/ (Date of access: 15.05.2019)..

Abovementioned factors lead to the conclusion that new media outlets are, according to the hypothesis of this research paper, less inclusive than traditional media that do listen to their opponents Transcript: Partial government shutdown causes delays in home loans [Electronic resource] // Fox News, 2019. URL: https://www.foxnews.com/transcript/partial-government-shutdown-causes-delays-in-home-loans (Date of access: 15.05.2019). and do not make news hooks out of statements belonging to people, who have next to nothing to do with politics. Data above confirms that despite being committed to a certain political ideology, old media do exercise a higher level of political pluralism by “neutrally” incorporating opposite side's opinion, while new media are obsessed with their opponents and/or certain figures among them.

3.1.3 Validating results (Data triangulation)

The dictionary that was used in the previoussubchapter needs verification, so a different dictionary provided by Hu and Liu will be used to re-examine data and confirm or reject findings that have been revealed above.

Before starting, it is crucial to remove two keywords that would otherwise undermine reliability of the outcome. Firstly, it is “a+” (positive) sentiment that must be removed due to misinterpretation by WordStat that considers it as an “a”plus everything after this letter: for instance, if this keyword is used in relation to a sentence “Political affairs between United States and China have not been easy lately”, it will highlight all “a” plus the part of the word that goes after the letter, resulting in “al”, “affairs”, “airs”, “Ates”, “and”, “a”, “ave”, “asy”, “ately” readout. Secondly, (positive) keyword (to) “trump” that is marked as a verb, not as a surname, needs to be ignored in the dictionary.

After that had been done, same operations were executed. Following results have been retrieved: firstly, proportion of positive and negative words in new media has changed, but not dramatically: the difference between these sentiments is now ~35% (Fig. 10,11).

Figures 10,11. Distribution of Positive and Negative sentiments(Dictionary 2, New Media)

Quite similar results can be observed in relation to negative and positive sentiments' rate per ten thousand words in new media with HuffPost and Breitbart accounting for the highest share of negative and positive sentiments respectively. Difference between them in the first case is ~116%, and ~109% in the second, which is slightly different from what was seen in the case when the main broad dictionary (Fig. 12,13).

Figures 12, 13. Negative and positive sentiments in new media. Rate per 10'000 words (Dictionary 2)

Moving on to traditional media outlets, results retrieved turned out to be different, but not to the extent that would undermine arguments mentioned above. Firstly, proportion of negative and positive keywords is ~1:1,03 which indicates a difference of only 3% (Fig. 14)

Figure 14. Distribution of negative and positive sentiments in old media (Frequency)

Despite that, proportion of negative sentiments stay generally the same with ABC falling behind the general mass and Fox News being between CBS and NBC. The only exception is that NBC and CBS have switched places with, again, an insignificant difference (Fig. 15). The situation changes again when it comes to positive sentiments, where Fox News showed a 62% higher share of positive keywords than its “left” counterpart of ABC (Fig. 16).This, while surely a big difference in absolute numbers, it is still almost twice as small as in the case with positive sentiments share in new media outlets. Moreover, the share of negative and positive sentiments per ten thousand words in Fox News as a separate entity indicates that their proportion is 1,095:1 respectively. This, along with other factors that have noticeably changed, still does not renderhypothesis insignificant, and confirms that the previous result is close enough to the current observations in terms of possible conclusions.

Figures 15,16. Negative/positive sentiments in traditional media. Rate per 10'000 words (Dictionary 2)

3.2 Qualitative content analysis

The previous subchapter has shown a clear pattern which indicates that old media are, while still being biased to a certain extent, exercise a higher level of tolerance towards their political opponents that are more widely included into the discussion, whereas new mediums either mock their opponents or ignore their presence at all while also tossing lots of negative/positive sentiments (depending on the side) into their articles. Hence, traditional outlets can be regarded as“battlegrounds” for political debate, while digital ones can be regarded as sides of the struggle

However, quantitative content analysis is not enough to say that this paper's hypothesis can be confirmed right away. That is why it is needed to apply to other methods of research, which will now be inductive qualitative content analysis that goes significantly beyond just counting negative sentiments and works with underlying meanings ofmessages delivered in articles and transcripts. IQCA

This subchapter suggests to examinein what way traditional and digital media outlets deal with their supposed political opponents in their articles. In order to determine their opponents, two rankings will be used throughout this part of the research - AllSides and MB/FC. The same random selection of two hundred articles will be used throughout this part of the research as well.

It should be primarily determined what are the “enemies” of each media outlet. According to AllSides and MB/FC, ABC, NBC, CBS and Politico lean left ABC News [Electronic resource] // AllSides, 2019. URL: https://www.allsides.com/news-source/abc-news-media-bias (Date of access: 15.05.2019); NBCNews.com [Electronic resource] // AllSides, 2019. URL: https://www.allsides.com/news-source/nbc-news-media-bias (Date of access: 15.05.2019); CBS News [Electronic resource] // AllSides, 2019. URL: https://www.allsides.com/news-source/cbs-news-media-bias (Date of access: 15.05.2019); Politico [Electronic resource] // AllSides, 2019. URL: https://www.allsides.com/news-source/politico-media-bias (Date of access: 15.05.2019)., Fox News leans right Fox News [Electronic resource] // AllSides, 2019. URL: https://www.allsides.com/news-source/fox-news-media-bias (Date of access: 15.05.2019). and Huffington Post with Breitbart are extremely left HuffPost [Electronic resource] // AllSides, 2019. URL: https://www.allsides.com/news-source/huffpost-media-bias (Date of access: 15.05.2019). and right Breitbart News [Electronic resource] // AllSides, 2019. URL: https://www.allsides.com/new s-source/breitbart (Date of access: 15.05.2019). respectively. As a result, it can be concluded that republicans are the main opponents of ABC, NBC, CBS, Politico and HuffPost, while Breitbart and Fox News' opponents are democrats and their supporters. Consequently,articles of each specific outlet must be analyzed concerning the medium's attitude towards their rival. Some of the findings were made possible with help from QDA Miner: encoding obscene/offensive vocabulary and arguments from authority into respective categories helped to detect some of the factors that pinpoint high levels of biasness in digital media.

3.2.1 Huffington Post, Breitbart and Politico

The first media outlet to be examined is Huffington Post that is considered to be one of the largest opponents of the current American republican establishment and Trump himself. Such attitude is illustrated by articles that claim Trump as a “violence encouragingLiebelson D., Stein S. Donald Trump Encourages Violence At His Rallies. His Fans Are Listening. [Electronic resource] // Huffington Post, 2016. URL: https://www.huffpost.com/entry/donald-trump-violence_n_56e1f16fe4b0b25c91815913 (Date of access: 15.05.2019)., lying Blumenthal P. Donald Trump Keeps Lying About New Jersey 9/11 Celebrations [Electronic resource] // Huffington Post, 2015. URL: https://www.huffpost.com/entry/donald-trump-911_n_565b1950e4b08e945feb7326 (Date of access: 15.05.2019). racist Marans D., O'Connor L. Here Are 13 Examples Of Donald Trump Being Racist [Electronic resource] // Huffington Post, 2016. URL: https://www.huffpost.com/entry/donald-trump-racist-examples_n_56d47177e4b03260bf777e83 (Date of access: 15.05.2019).”. Considering the government shutdown, HuffPost has mentioned republicans or Trump in almost every article devoted to this case, but at the same time has never applied to their speeches or statements in a neutral, unbiased, way. It goes without saying that Huffington cited their opponents multiple times, but these citations were aimed at defacing them later in the text. For instance, the case when Trump's adviser Kevin Hasset said that government shutdown had some pros for employees that had an opportunity to go on a vacation that did not affect their days off count. His quote goes as follows: “A huge share of government workers were going to take vacation days, say, between Christmas and New Year's, and then we have a shutdown, and so they can't go to work, and so then they have the vacation, but they don't have to use their vacation days. And then they come back, and then they get their back pay… Then they're, in some sense, they're better off Visser N. Trump Adviser Suggests Unpaid Government Workers Are `Better Off' During Shutdown [Electronic resource] // Huffington Post, 2019. URL: https://www.huffpost.com/entry/kevin-hassett-government-shutdown_n_5c3be402e4b01c93e00b0640 (Date of access: 15.05.2019).”. Huffington rushed to name their article “Trump Adviser Suggests Unpaid Government Workers Are `Better Off' During Shutdown” Ibid. ignoring his original quote and subsequently creating a straight-off negative attitude towards that person among the readers. It is also worth noting that Breitbart completely ignored this case Kevin Hassett [Electronic resource] // Breitbart, 2019. URL: https://www.breitbart.com/tag/kevin-hassett/ (Date of access: 15.05.2019)., whereas Fox News, that would presumably do the same, did not Trump's economic adviser compares shutdown to vacation, says furloughed workers are `better off' [Electronic resource] // Fox 43, 2019. URL: https://fox43.com/2019/01/13/trumps-economic-adviser-compares-shutdown-to-vacation-says-furloughed-workers-are-better-off/ (Date of access: 15.05.2019)..

It is also worth noting that Huffington Post writers pay a lot of attention to celebrities and comedy shows that publically condemn republicans' actions and blame the current American president for all the issues they are having. For example, one of the entries contained by the sample for quantitative content analysis was solely devoted to a comedian Jimmy Kimmel who said Trump's “one crazy zigzag stroke…” could let him “… be back on the greens at Mar-a-Lago faster than you [Trump] can say `Pocahontas' Moran L. Kimmel Makes The Most Persuasive Case For Trump To End Government Shutdown [Electronic resource] // Huffington Post, 2019. URL: https://www.huffpost.com/entry/jimmy-kimmel-donald-trump-shutdown-case_n_5c403cb9e4b041e98ffac375 (Date of access: 15.05.2019).” relating to president's absence from golfclub that had lasted for 52 days. It goes without saying that both Breitbart, along with Fox News and pretty much every other media outlet in scope ignored this message. The other similar case was with a famous rapper Cardi B who was mentioned in four out of 34 Huffington selected articles. Cardi B “blasted Pres. Trump over government shutdown in viral video Moran L. CNN's Don Lemon Likens Government Shutdown To Slavery [Electronic resource] // Huffington Post, 2019. URL: https://www.huffpost.com/entry/don-lemon-donald-trump-slavery-shutdown_n_5c41c7f0e4b0bfa693c2aad7 (Date of access: 15.05.2019).”, “returned to social media… to share her frustration with President Donald Trump Richards K. Cardi B Has Some Advice For Trump Supporters During The Government Shutdown [Electronic resource] // Huffington Post, 2019. URL: https://www.huffpost.com/entry/cardi-b-government-shutdown-trump-supporters_n_5c439883e4b0a8dbe171f19a (Date of access: 15.05.2019).” Moreover, one of HP's article is titled “Cardi B's Trashing of Donald Trump For The Shutdown is Now a Banging Song Moran L. Cardi B's Trashing Of Donald Trump For The Shutdown Is Now A Banging Song [Electronic resource] // Huffington Post, 2019. URL: https://www.huffpost.com/entry/cardi-b-shutdown-donald-trump-remix-song_n_5c41976ce4b0a8dbe16fdb09 (Date of access: 15.05.2019).”. Behavior demonstrated above was ignored by major traditional media outlets and Politico as well, but was squeezed out by Breitbart. This medium utilized a different approach which tried to persuade readers that the artist is not the person to be listened to. They cited her tweets with plenty of obscene language and offenses towards Trump supporters Caruso J. Watch: Rapper Cardi B Targets Trump in Endorsement of Bernie Sanders for President [Electronic resource] // Breitbart, 2019. URL: https://www.breitbart.com/entertainment/2019/04/15/watch-rapper-cardi-b-targets-trump-in-endorsement-of-bernie-sanders-for-president/ (Date of access: 15.05.2019) with also reminding the audience that Cardi B “recently admitted to drugging and robbing men while she was younger Ibid. ”. The artist was mentioned by Breitbart two more times (across the gathered articles). One of the articles that was, again, devoted to her, Breitbart editors reported that “Cardi B has been invited to a Democratic event in Iowa after calling supporters of President Trump “f****** racist rednecks Caruso J. Cardi B Invited to Iowa Democart Event After Calling Trump Supporters `F*cking Racist Rednecks' [Electronic resource] // Breitbart, 2019. URL: https://www.breitbart.com/entertainment/2019/01/30/cardi-b-invited-to-iowa-democrat-event-after-calling-trump-supporters-fcking-racist-rednecks/ (Date of access: 15.05.2019). ” and repeated the same quote in an article that covered criticism of Donald Trump by Cardi's husband, rapper Offset who also expressed his discontent by saying “that's some slave s***” Caruso J. Rapper Offset Melts Over Government Shutdown: `That's Some Slave Sh*t' [Electronic resource] // Breitbart, 2019. URL: https://www.breitbart.com/entertainment/2019/01/23/offset-blasts-government-shutdown-thats-some-slave-sht/ (Date of access: 15.05.2019).. Plus, BB cited and “debunked” a “far-left” pop star Cher who stated that FBI must “Stick Trump in Straight Jacket and Put His A** in Prison Cher: Feds Should `Stick Trump in Straight Jacket and `Put His A**' in Prison [Electronic resource] // Breitbart, 2019. URL: https://www.breitbart.com/entertainment/2019/01/24/cher-feds-should-stick-trump-in-straight-jacket-and-put-his-a-in-prison/ (Date of access: 15.05.2019).” (this is how the article is titled). It thus can be concluded that HuffPost, along with Breitbart, often cite artists, not real politicians and do it one-sidedly: in the first case it is done to show discontent with Trump's policies among celebrities, and the second one'sjournalists try to deface democrats and liberals by choosing stars who can't choose words when criticizing someone.