DEV Community

Cover image for Media Effect on BLM: A Sentiment Analysis on News Media Outlets’ Reaffirming Narratives

Posted on

Media Effect on BLM: A Sentiment Analysis on News Media Outlets’ Reaffirming Narratives

On the night of February 26, 2012, Trayvon Martin, a 17-year-old black male, was walking home when he was shot and killed by George Zimmerman. Trayvon Martin was not the first innocent African American boy to face an unlawful killing. Nevertheless, his death was a significant steppingstone towards bringing the community together to birth the Black Lives Matter movement. Trayvon's case became a media feeding frenzy among civil rights activists, political pundits, politicians, and even gained the attention of then President Barack Obama, who made a public address on the killing. In 2013, three radical Black organizers — Alicia Garza, Patrisse Cullors, and Opal Tometi — created a Black-centered political will and movement building project called #BlackLivesMatter. The group was created in response to Trayvon Martin's murderer, George Zimmerman, being acquitted of all charges (BLM, 2020).

Black Lives Matter is an ideological and political intervention in a world where Black lives are systematically and intentionally targeted for demise (BLM, 2020). It is an affirmation of Black folks' humanity, our contributions to this society, and our resilience in the face of deadly oppression (BLM,2020). The Black Lives Matter movement has steadily gained support over the years; the BLM movement now has over 40 chapters around the United States. Even with the increased growth of the BLM movement, media coverage has been declining. Dhrumil Mehta of FiveThirtyEight examined the popularity of the BLM movement in mass media:
"The national media's attention has shifted to the protests against police brutality that have arisen across the country following the death of George Floyd, a black man killed by a police officer in Minneapolis. However, that brutality and those protests are nothing new — on-duty police officers have fatally shot about 1,000 Americans every year between 2015 and 2019. Black people are consistently most at risk of being killed by police. This year has been no different so far. However, the media has not paid much attention to protests against police brutality or misconduct over the last few years. These protests, often grouped under the Black Lives Matter movement's umbrella, featured prominently in national media during and after the 2014 demonstrations in Ferguson, Missouri. In recent years, though, they have received much less media attention (Mehta, 2020).

Mehta illustrates a common problem with coverage of the protests. Mainstream media outlets focus on the act of protesting rather than the content of the protest message. The media has been historically concerned with making money, and when a topic is no longer generating revenue because it is considered average news, they stop covering it to protect the bottom line.

National media coverage of BLM protests saw a decline between 2015 and 2019. With the deaths of George Floyd and Breonna Taylor in 2020, BLM has gotten a renewed surge of media coverage. With this surge has come increased animosity toward the movement as well. Counter framing and anti-narratives of BLM have created counter-protest movements such as All Lives Matter and Blue Lives Matter. The connection between the media narrative and the increase of counter-protests has led the researcher to ask how media depiction of BLM shapes public opinion on protests.

Media Framing

Framing theory has its roots in many schools of thought (Littlejohn, Foss, and Oetzel, 2017). Each school of thought has its way of defining media framing. Bateson (1972) first posited the concept of framing when he defined psychological frames as a "spatial and temporary bonding of a set of interactive messages" (p. 197) that operates as a form of metacommunication (Hallahan, 2008). Framing describes the practice of thinking about news items and story content within a familiar context. Media framing is a tool where media engagement influences people's choices about how to process given information. "Framing theory explains that the media create these frames by introducing news items with predefined and narrow contextualization. Frames can be designed to enhance understanding or are used as cognitive shortcuts to link stories to the bigger picture" (Arowolo, 2017).

When discussing the framing process, Gamson and Modigliani (1987) described packages that elites and media use to characterize an issue. These packages are composed of arguments, information, symbols, metaphors, and images. These packages affect how people understand, interpret, and react to a problem or issue. Tewksbury and Scheufele (1999) build on Gamson and Modigliani's information theory frame, arguing that information can affect audience members' beliefs about the issue and its treatment as well. Framing unites information into packages of messaging; framing is the glue that binds information together to represent an overarching theme. These themes then influence the audience's ability to process information.

Protest Paradigm

Social activist and social movement groups are in a constant battle regarding media representation. These groups struggle to gain accurate news representations; mainstream news media rely on dramatic interpretations of news to attract viewership and subsequently advertisement revenue (Chan and Lee, 1984; McLeod and Hertog, 1999; Gamson and Wolfsfeld, 1993; Kilgo and Harlow, 2019). This drive for profit results is a reductive representation of activist groups. These less than favorable representations tend to focus on more dramatic storylines instead of activist demands and successes (Brasted, 2005). Even with news media’s biased representation, social activist groups still rely on mainstream media to get their name and mission out to the public.

With the growth of other forms of news such as social media and internet blogs, social activist groups still find themselves at the mercy of mainstream media (Roskos-Ewoldsen, 2007). Activist groups utilized these social platform mediums to project their goals and network(broadcast) their messages to broader audiences. Journalists and other media sources have routinely found little news value in demands for change without action, so protestors create media interest by employing action (Bennett & Segerberg, 2012). With increased activity comes increased media coverage, but this media coverage is not always positive or accurate.
Mainstream media may not be the only channel for news consumption, but it is the channel that provides legitimacy. Because social activist groups have a need for mainstream legitimacy in order to promote their demands for change, activist groups are always in a dilemma. This phenomenon is referred to as the protest paradigm, originally coined by Chan and Lee in 1984. Chan and Lee developed the term in their study which described the relationship between media outlets and Hong Kong protest groups. Chan and Lee (1984) voiced the concept of the protest paradigm to refer to the pattern of delegitimizing news coverage of protest and dissent.

Building on Chan and Lee’s concepts, McLeod and Hertog (1999) found repeating patterns that took place in media representation of activist groups. McLeod and Hertog defined these recuring elements as a “routinized pattern or implicit template for the coverage of social protests. These patterns, known as the protest paradigm, are ultimately created by the symbiotic but imbalanced relationship between activists and the press (Kilgo and Harlow, 2019). This relationship develops in the following way: (1) the press does not cover movements that do not engage in newsworthy activity; (2) advocates stage events to attract media attention, which they need to distribute their message to broader audiences and to signal their strength; and (3) journalists then cover these staged events without generating substantive information about the event’s background or the grievances or agendas of the movement behind the protest. This paradigm thus is a paradoxical situation: Movements must adapt and appeal to media logic to receive coverage, but when they do, protest coverage tends to demonize protesters, characterizing them as menaces to society, marginalizing their voices, and under or inadequately reporting the grievances, demands, and agendas of movements (Chan & Lee, 1984; Kilgo and Harlow, 2019; McLeod and Hertog,1999). These elements come together to shape news representation of activism groups; Kilgo and Harlow’s research can be used as a method to better apply Chan and Lee’s concepts on current news media coverage. Chan and Lee’s, together with Kilgo and Harlow’s, concepts are still currently applied in contemporary news media coverage; however, researchers have noted that the contemporary news media can exhibit a more complex relation to the politics of protest than assumed in the past (Cottle, 2008, p. 859). This indicates that there is a need to update the approaches we take in examining the relationship between media representation and social movement groups.

Studies focusing on the concept of the protest paradigm have also begun to treat the pattern of media coverage as a variable—that is, instead of assuming or trying to prove that the mainstream media are biased against social protests, the pattern of delegitimizing coverage is treated as existing to varying degrees in the coverage of different types of protests, by different media, or in different types of societies (Lee, 2014). Boyle et al. (2012) found that protests employing more radical tactics are portrayed more negatively. McCluskey, Stein, Boyle, and McLeod (2009) found that news coverage of protests in U.S. newspapers belonging to less diverse communities conforms more to the protest paradigm. This is because less diverse communities do not have established mechanisms for resolving conflicts; thus, the news media organizations have a stronger role in maintaining and creating community consensus.
Media Framing of Black Lives Matter

The Black Lives Matter movement, like other social movements, relies on mass media to gain attention. Black Lives Matter relied on both internet media sources and conventional mass media for audience engagement minorities (Roskos-Ewoldsen, Klinger, & Roskos-Ewoldsen, 2007). They utilized these mediums to project their goals and network their messages to broader audiences. Journalists and other media sources have routinely found little news value in demands for change without action, so protestors create media interest by employing action (Bennett & Segerberg, 2012). With increased activity comes increased media coverage, but this media coverage is not always positive. The BLM movement suffers from the protest paradigm; this paradigm refers to conventional media focusing on drama, violence, and confrontation rather than details such as protest agendas, demands, and accomplishments (Brasted, 2005). Detenber and Gotlieb (2007) found highly critical, and high conflict media articles, made participants more skeptical of protests than those that were exposed to noncritical coverage. Detenber, Gotlieb, McLeod, and Malinkina (2007) found that well-established beliefs were essential to consider when assessing media effects and attitudes toward protest. They found that preexisting attitudes and beliefs were fundamental components of protest image interpretation. Mass media can use these preexisting beliefs to push a protest narrative to be more favorable to their audience. Therefore, the media depiction of protest ranges widely between news networks, as each news network frames their coverage to affirm their viewership's beliefs and attitudes.

As media viewership has become more polarized, political ties and media sources have contributed to the audience's selective exposure. Selective exposure theory refers to how audiences choose media that reinforces preexisting views while avoiding contradictory information (Lanier, Frush, and Holt, 2019; Dustin and Carnahan, 2020).
Studies on race in the media find that racial minorities are among the most clearly defined out-groups (Purdie-Vaughns et al.,2008). People tend to view out-group members negatively, and audience members often evaluate media that can either confirm or dis-confirm their preexisting biases (Vescio & Biernat, 1999). Selective exposure theory tries to explain the connection between race and selected media choices. Selective exposure tries to explain that people are motivated to find news that affirms their racial group's perceived specialness and portrays their group in a most favorable light (Appiah, 2013). "Selective exposure analyses also rarely explain the challenge faced by African Americans when seeking positive news stories to boost their group identity. Media rarely breaks from their routines and often cover news from the dominant group (Bjornstrom, E., Kaufman, R., Peterson, R., Slater, M., 2010). Because people of color are always looking at media from an outsider's perspective, the media effect analysis on news media should examine the dominant groups' reaffirming narrative. This reaffirmation creates an echo chamber of thoughts that leads to many negative racial stereotypes in the news. An examination of these dominant echo chambers could lead to a new understanding of the effect viewership of specific news media has on minority groups' perspectives within the United States. This paper proposes analyzing popular news media on opposite sides of political leaning. Through this examination, this paper can illuminate the framing and priming that goes into polarized political news media.

Research Questions and Hypotheses

RQ1: Does the audience’s political leaning alter the media’s framing and utilization of priming on the Black Lives Matter protest?

This paper theorizes that politically polarized media audiences will significantly affect the amount of negative or positive framing and priming within a newscast of the Black Lives Matter protests. Under the frame of the protest paradigm, news media will focus on the entertainment value of a movement. Meaning that news media framing of protests will lean towards entertaining their viewership (Berkowitz, 1986; Linus Abraham & Osei Appiah, 2006). This paper theorizes that political polarization alters the types of audiences that news media receive and subsequently what types of framing and priming utilization the media channels use towards appeasing audience members.

Typically, a conservative news media outlet will have high Republican viewership, whereas liberal media has a high Democrat viewership. Because Black Lives Matter has had more ongoing support from Democrats it is logical to theorize that the media's framing and priming will lean towards the democrat audience members.

Sentiment Analysis

This paper utilized a computer-assisted sentiment analysis of BLM news headlines. Sentiment analysis (also known as opinion mining) is a natural language (NLP) technique used to determine whether data is positive, negative, or neutral. Sentiment analysis has been used as a method for text analysis and computational linguistics to interpret information from a given source. It has successfully been used on social networks to extract useful information specially for instance for customer services and marketing concerns (Niklander and Niklander, 2017).

Niklander and Niklander used a combination of sentiment analysis and content analysis to investigate Twitter tweets surrounding the #NiUnaMenos (Not One Less) social movement. In their research, they used a SeniStrength. SeniStrength is a tool that can perform sentiment analysis. This tool is predominantly used for the analysis of social media. SentiStrength was not used in this study because it heavily relies on a word bank with prescribed values. After examining the words within the word bank, I found that it would not give a proper analysis of news headlines.

*Research Tools

For this research, I created a python script that downloaded news headlines surrounding BLM. It then used TextBlob and VADER to do sentiment analysis on those news headlines. TextBlob sentiment analyzer returns two properties for a given input sentence. Polarity which is a float variable that lies between [-1,1], -1 indicates negative sentiment and +1 indicates positive sentiments. Subjectivity is also represented as a float variable that lies in the range of [0,1]. Subjective sentences generally refer to opinion, emotion, or judgment.
VADER uses a list of lexical features (e.g., words) which are labeled as positive or negative according to their semantic orientation to calculate the text sentiment. Vader sentiment returns the probability of a given input sentence to be positive, negative, and neutral. A sentiment score is then generated ranging from -1 to 1, where -1 is for negative sentiment, 0 is neutral and +1 is a positive sentiment. These scores are then combined into a single score called a compound score. This score gives the average sentiment of the sentence.

News Selection

News websites were selected based on their popularity. For this study, it was important to get a wide range of news websites. I used Semrush's most visited news website chart to determine what news websites to focus on. From their analysis, I chose the top news website from every state in the United States.

I chose to set no parameters (such as date) on the news searches. Google algorithm returns the most visited website and websites that match the keywords the most. Therefore, whatever headlines are returned are those that are being viewed the most by the general population. This would give the most accurate variable when examining the implications of the news headlines. I took over 300+ headlines from each news platform.


After collecting and cleaning the data, I took the average of each news website and placed them with their corresponding state. The results of the study are shown below.

VADER, polarity, and subjective scores are represented in a range of positive (1) and negative (-1) where 0 is neutral.

After gathering the results for each state, I then put them into groups based on political leaning. I used the World Population Reviews 2022 Political Party to determine which states should be combined. The political parties were categorized into three political parties (Democrat, Republican, and Split Party). An average score for each party was created.

Democrat states scored an average of -5.812% VADER, 32.895% Polarity, and -5.733% Subjective. Indicating that news on BLM within these states had an average of -5.8% tilt meaning that news was slightly negative. These states also have a 32% Polarity rate indicating that 32% of news within these states had a positive tilt. These states also have a -5.7% subjectivity score, indicating that news in these states are slightly opinion-based.
Republican states scored an average of -5.625% VADER, 31.789% Polarity, and -6.095% Subjective. Indicating that news on BLM within these states had an average of -5.6% tilt meaning that news was slightly negative. These states also have a 31.7% Polarity rate indicating that 31.7% of news within these states had a positive tilt. These states also have a -6.09% subjectivity score, indicating that news in these states are slightly opinion-based.
Split states scored an average of -6.511% VADER, 27.129% Polarity, and -8.155% Subjective. Indicating that news on BLM within these states had an average of -6.5% tilt meaning that news was slightly negative. These states also have a 27.1% Polarity rate indicating that 27.1% of news within these states had a positive tilt. These states also have a -8.15% subjectivity score, indicating that news in these states are slightly opinion-based.

Based on these results the hypothesis was found to be correct. When looking at Democrat states there was less opinionated news towards BLM when compared to Republican states. There was also more positive news regarding BLM in Democrat states. Democrat states had a slightly higher negative VADER score. But because of the Polarity and Subjective score, I believe that this can be seen as an anomaly. When comparing all three parties the results indicate that split states had the highest level of negative bias towards BLM. Indicating that there is a connection between state parties and news representation of BLM. States with high levels of Democrats had a less negative view of BLM within the news they consume, whereas states with high levels of Republicans had higher negative views of BLM within the news they consume. Split states indicate that there is a high level of competing understandings of BLM within the news. Split states have a wide range of online news sources that they consume, creating a wide range of news representation technics when discussing BLM.


The researcher found political leaning altered media framing and utilization of priming on the Black Lives Matter protest. News within Republican states framed and utilized media priming to be consistent with their viewership. This conservative viewership has been shown to agree with violent depictions of minority protesters. Conservative viewership has already established beliefs towards protesters and law enforcement. Conservatives believe in law enforcement and believe that instances of police brutality are not expected.
Democrat state news had a less harmful framing of the protest but still utilized strong framing and priming forums. Though these forums were less negative, they can still be seen as tilted in bias toward some conservative viewers. Democrat states have strong liberal viewership; thus, it was unsurprising to find that news would be framed more positively reflecting the liberal viewers of BLM.
The results show that the political leanings of viewers affect the way news media frames protest movements. The protest paradigm argues that media frames protests to be consistent with the viewer's belief systems. This type of framing has a few implications that should be addressed.

*Implications *

Because news media appeases its viewership, framing and utilizing media priming is used to push an agenda. This type of agenda creates polarization. Selective exposure theory argues that viewers gravitate to media that aligns with their beliefs. This reaffirmation creates an echo chamber, resulting in increased nodes. These nodes create an opportune environment for news media to continue to prime narratives. These negative narratives can result in radicalization on either side. This radicalization can lead to negative actions toward protesters. This type of radicalization can be seen in Kenosha, Wisconsin, where Kyle Rittenhouse, a 17-year-old boy, fired on protesters with a military-style semi-automatic rifle, killing two and seriously injuring a third (BBC, 2020). This paper does not argue that this violent incident resulted from political leaning, but that this polarization through tilted news media viewing could have a similar effect on viewers' attitudes and actions.

*Study Limitation/Future Research *

The researcher examined online news media from every state in America. There was a significant amount of data that was collected about 300+ news headlines for each news platform. Even with these numbers, I believe that there is a need for more data. A study that collected new headlines from the past decade would better demonstrate a protest paradigm variable.

There was also a limitation to the sentiment analysis programs that were used. Both tools are based on dictionaries that do not properly reflect the words that are found in news headlines. To better tackle this problem, I believe that there will need to be a system built from the ground up that has a dictionary created from the most common words used in news headlines. This new program will also need to create a set of rules for dealing with phrases and must be able to determine sentiment based on BLM being the object of a sentence of the subject. I plan to continue this research and create a tool that can be used to better explain the protest paradigm, through this new program researchers can treat the protest paradigm as a variable that can be used to do correlation analysis. This would provide a new lens to research and to try and break down the bias that happens in news media representations of social activism groups.

*Conclusion *

As previously stated, people of color are always looking at media from an outsider's perspective, so a media effect analysis on news media should examine the dominant group in reaffirming the narrative. The researcher examined the news media they believed would create those echo chambers. The research found that media framing and priming were consistent with the hypothesis. News media companies created media framing that supports their audience's attitudes and beliefs. Media-created echo chambers are a result of selective exposure theory. This creates a feedback loop where news media outlets create and reaffirm belief systems. This happens via the creation and re-energization of nodes. Through this node feedback, news media has an unsurprising effect on viewers' attitudes and actions towards protesters.

There needs to be more research in this area to better break down the impact this type of framing can have. This research showed that there is already a connection between biased news representation and state political parties. Furthering the field by creating tools that can be easily used by other communication scholars is a goal of mine, and something I believe is needed to further research past the ideas of proving the protest paradigm and into research that views and treats it as a variable.

Top comments (0)