Posted on / by Saikiran Kannan / in Long Story

Grappling With Bias

Indian news’ unbalanced coverage of the ongoing crisis within the Wrestling Federation of India, characterized by skewed narratives and selective reporting, might hinder India’s chance to create a pivotal moment similar to the Larry Nassar case.

The Larry Nassar sexual abuse case received significant media coverage in the United States. Nassar was a former USA Gymnastics national team doctor and Michigan State University physician accused of sexually abusing hundreds of young female athletes under the guise of medical treatment. Larry was convicted, and this brought about wholesale changes to how America’s sporting bodies function and act when dealing with allegations of sexual harassment. 

Now, India is facing a similarly precarious situation where its Olympic medalists are accusing the Wrestling Federation of India president (who has temporarily resigned) of sexual harassment and abuse of power.

The facts of the case are that the accused, Brij Bhushan Sharan Singh—a member of Parliament belonging to India’s ruling political party, the BJP—has refuted all charges and called the protests by female wrestlers a conspiracy to remove him from office. 

Moreover, the Delhi police—controlled by the central government led by the BJP—didn’t file FIRs against the accused until the Supreme Court of India intervened. This prolonged inaction, despite the allegations being cognizable and mandating immediate registration of FIR by the police as per Indian law, turns the tables of “political conspiracy”.

While the law will take its course, we look at how Indian news media has been covering this amidst all the other distractions, be it the Indian Premier League, the elections in the state of Karnataka, or the new parliament building inauguration.

Contextualizing Bias

The analysis period of news articles was from January 18, 2023, to May 9, 2023. 350+ articles from 50+ media houses were used to analyze reporting patterns. The articles were primarily analyzed to identify two things:

  1. If they got too political or apolitical
  2. If they focused enough on the allegations of sexual assault

Starting with a keyword analysis and keeping the context of the issue as a basis, we quantitatively evaluated all the articles using terms like BJP, Modi, Government, MP, Lok Sabha, etc. Once we logged the occurrences of these terms for each piece, we then used the overall article lengths and the proportion of the keywords in them to arrive at a summarized view of the keyword density ratio for each publisher. We list the top ten in the image below.

Hub News, Frontline, The Asian Age, News Click, Siasat.com, The Logical Indian, DNA India, NDTV.com, The Federal News, and Outlook India occupy the top ten spots. 

This ranking essentially refers to articles with a high degree of mention of the political affiliations of the accused, the ruling party, and their statements relating to this topic. 

There needs to be a sense of thoughtfulness in news coverage owing to the issue’s sensitivity. As this matter involves allegations of sexual assault, it is essential that while the general tendency of reporting will focus on the politics surrounding the incident, the core interest of justice for the affected athletes and the aspects of sexual harassment does not get buried. 

To draw parallels to the coverage of a similar nature, Journalist Marta Arráez, while covering the sexual assault allegations and proceedings against an ex-Barcelona footballer in Brazil, expressed concerns about how such cases get reported by the media. Back then, she particularly accused certain media outlets of focusing on the consequences that the footballer would face in case of being found guilty rather than addressing the story from the alleged victim’s perspective. 

Keeping this in mind, the next set of keywords we used to analyze the articles were Harassment, Women, Female, Sexual harassment, etc., so as to keep the contextual perspective in mind. We list the top ten in the chart below.

The top ten spots are occupied by Wion, Rising Kashmir, Mid-Day, Hub News, Bar and Bench, Indiatimes.com, The Economic Times, The Statesman, The Federal, and Zee News. 

We examined a few articles individually to see how the writing patterns varied. For instance, a Frontline article focuses on the political association of the chief accused, Brij Bhushan Singh, who was also the President of the WFI (Wrestling Federation of India) when the allegations were first brought to light. It also does not use terms like Women wrestlers or sexual harassment. 

On the other hand, Wion never associates Brij Bhushan Singh with the BJP and only refers to him as the WFI chief. WION, however, constantly refers to allegations of sexual harassment and the women wrestlers involved in the protests. In a Republic story, part of the analysis also follows a similar pattern of not referring to the political affiliations of Brij Bhushan but referring to female wrestlers and allegations of sexual harassment.

Some media houses are tall on both fronts regarding the degree of political notations in their stories and references to the sexual harassment angle. An example of this is Hub News

The above image shows how it offsets both writing patterns concerning covering the protests. 

Aside from the keyword analysis, there is also a matter of contextual analysis of the articles to ascertain the core topic a media houses focus on. The below matrix shows media houses like Frontline (part of the Hindu Group), Times of India, News Click, Scroll, ABP Live, etc., focused more on political notations and the politics in this issue rather than the sexual harassment angle. The media groups in this bucket could be both left and right regarding political affiliations or bias. 

On the other hand, the media organizations that focused more on the sexual harassment angle and offered more context-specific reporting and sensitivity are listed in the below matrix. Wion, Rising Kashmir, The Wire, Deccan Herald, India Today, etc., ranked highest in this ranking.

Contextual and consistent reporting by media houses is vital for the case to move toward a logical and just conclusion.

For example, in the Larry Nassar case, American news followed the story for years, ensuring it stayed alive, gained traction, and enabled more victims to share their experiences. They reported on the testimonies of Nassar’s victims, interviews with survivors, legal proceedings, and reactions from the gymnastics community and public figures.

Hence, the Larry Nassar case garnered attention not only for the shocking nature of the allegations but also because it exposed systemic failures within USA Gymnastics and Michigan State University, as many of the victims had reported the abuse to authorities but were not taken seriously.

“Frontline (part of the Hindu Group), Times of India, News Click, Scroll, ABP Live, etc. were reporting more on the politics of the issue rather than the issue itself, while Wion, Rising Kashmir, The Wire, Deccan Herald, India Today, etc. reported more on the Sexual harassment angles”

Reader-friendly Reporting

In terms of analyzing the reader-friendliness of the articles, we looked at the below attributes to normalize a final score and rank the media agencies. 

Firstly, we looked at the article lengths across the analyzed data set. The article lengths are anywhere between a few hundred to 3000 words. On average, each article is about 500 words. To understand the reader-friendliness of the media agencies via their articles, we conduct the “Vocabulary density analysis”.

It refers to examining and measuring the richness and diversity of vocabulary used in a piece of text. It involves quantifying the variety of words and their frequency of occurrence within the text. The analysis assesses the text’s complexity, sophistication, and lexical diversity.

Looking at the mean article lengths, we can see that Newslaundry, Zee Business, Sportstar, Khel Now, and IndiaTimes.com occupy the top five spots. All these have mean article lengths above 1000 words per article.

Using the findings to plot an article length distribution chart shows that the maximum number of articles is between 500 and 750 words.

The next step was to perform an article vocabulary analysis. The chart below ranks the top 20 news agencies in terms of the number of unique words used in each article.

News Laundry, followed by The Asian Age and Sportstar, have the highest lexical richness and hold the top three spots. One cannot ignore the fact that the mean article lengths and the lexical richness are correlated; hence after normalizing the findings, we get a different picture.

The above graph shows the ranking of media publishers having high lexical richness per word after normalization based on the respective length of the articles.

Considering all the above results, we can summarize and rank the top ten media houses on their reader-friendliness score. Do note that in the below graph, the lower the score, the higher the reader-friendliness of the media house.

We can note that these ten media houses have the maximum reader-friendliness among our datasets. This overall score considers the article lengths, the ratio of unique words, and the ratio of sophisticated vocabulary used in the articles. We can summarize that the lower the sophistication and unique words, the shorter the article lengths, and the better the reader-friendliness. 

“Zee Business, PTI, Khel Now, Indiatimes, Sportstar, Hub News, Newslaundry, OpIndia, The Asian Age, etc. had the best reader-friendliness among all the publications reporting on this topic”

Reader-friendly media coverage is important to spark broader discussions about sexual abuse, accountability, and the need for reforms in the world of competitive sports, as it will reach more people in an easily understandable format. The impactful coverage of Larry Nassar’s case in the US emphasized the bravery of the survivors in speaking out against Nassar and shed light on the institutional failures that allowed him to continue his abuse for so long. It remains to be seen if the Indian media can play a similar role in unmasking the truth in allegations leveled against Brij Bhushan Singh and if they help in the case reach a logical conclusion.

Emotional Flow, Sentiment Analysis, and Objective Reporting

In terms of a wholesome analysis of the emotional flow and sentimental analysis of articles (each article is broken down into multiple blocks of 200 words each and then analyzed and bucketed into either anger, fear, joy, or sadness, and then a mean value of the sentimental context is identified), most articles exhibit anger, and almost no article exhibited sadness. 

The below image shows the overall sentimental analysis of all the media organizations whose articles are part of our dataset. Here, the “red” refers to anger, “cyan” refers to fear, “green” refers to joy, and “yellow” refers to sadness.

Media houses like ESPN, Hub News, Mid-Day, NDTV, Newslaundry, OpIndia, Outlook, PTI, Republic, SwarajyaMag, The Week, The Wire, The Print, etc., displayed the maximum anger in their articles. Of course, the degrees were high, but the polarities could be on opposite sides. 

Looking at polarities, it is important to also look at the degree of objectivity in articles as part of our analysis. In our next analysis, we looked at the correlation between the subjectivity of an article and the polarity of the same. We can see from the below chart that the higher the polarity, the higher the subjectivity or opinionated articles from the respective media houses. 

The positive polarities refer to the density of positive words in the articles, whereas the negative polarities refer to the density of negative words in the articles. We can understand them as the overall sentiments of the articles split between positive and negative.

Media like DNA, Mid-Day, PTI, etc., seem to have the least subjectivity in their articles but are still placed in varying planes of polarities. We see a 50% to 60% correlation between subjectivity and polarity, even with outliers.

We tried to identify three major clusters from the dataset based on the extracted features from the articles. The number of articles and the overall dataset falls below the minimum threshold required; hence, clustering the media houses was challenging. Using our pre-trained models, We combined the articles’ emotional flow and sentimental tonality with contextual understanding. As per statistical norms, the clusters can be understood differently depending on how a reader interprets the findings.

“Media houses like ESPN, Hub News, Mid-day, NDTV, Newslaundry, OpIndia, Outlook, PTI, Republic, SwarajyaMag, The Week, The Wire, The Print, etc. displayed the maximum degree of Anger in their articles despite being on opposite sides of the political spectrum”

We have a total of three clusters. Red-colored pointers identify cluster 1. Cyan-colored pointers identify cluster 2. Green-colored pointers identify cluster 3.

There is a good chance the same media house may belong to multiple clusters, so it’s useful to identify their dominant cluster to understand a publisher’s reporting patterns. Of these, Cluster 1 and 2 amount to 82 per cent of all the media articles; hence, we will focus more on them.

At a high level, Cluster 1 (Red) can be understood as having a slight positive bias meaning having a more contextually positive tone. Cluster 2 (Cyan) can be understood as one with a slight negative bias, meaning having a more contextually positive tone, and Cluster 3 (Green) can be deemed neutral tonality.

To simplify the reading of the above chart, Wion, Bar & Bench, DNA, News18, etc., can be seen as having the most neutral tone of the lot. Red and Cyan cluster-dominated media houses will have biases on either side of the polarities. 

Clusters 1 and 2 can be understood better by looking at the below analysis, where we look at word clouds and lexical dispersions, respectively.

Cluster 1: 

This cluster is less about the political affiliations and more about the allegations and the chief accused.

Cluster 2: 

This cluster is more about the political affiliations and the wrestler protestors. 

Cluster 3: 

This cluster concerns the chief accused, the protests surrounding the allegations, and negligible mention of the political affiliations. Only 18 per cent of the articles fall under this cluster. 


Concluding Thoughts

The persistent and objective reporting by the US media played a crucial role in the Larry Nassar sexual abuse case, leading to significant outcomes. Indian news can learn from this, adopting a balanced approach of objectivity and sensitivity when covering similar stories.

Moreover, the dogged media coverage of the Nassar case profoundly impacted various aspects of society. Survivors were given a platform to share their stories, leading to increased empathy, support, and solidarity. The extensive coverage raised awareness about the prevalence of sexual abuse in sports and emphasized the need for stronger safeguards and accountability measures.

Most importantly, the media scrutiny exposed institutional failures within USA Gymnastics and Michigan State University, holding them accountable and promoting policies, leadership, and athlete welfare culture reforms.

The coverage inspired other survivors to come forward, initiating a broader conversation about sexual abuse and fostering a culture of accountability.

The public pressure the coverage raised in favor of legal action against Nassar led to the introduction of legislation to protect athletes and prevent future abuse. It even formed the basis for a cultural shift, challenging societal norms, promoting discussions on consent and boundaries, and emphasizing the importance of believing survivors.

While the jury is still out on both the case against Brij Bhushan Singh and the Indian media’s coverage of it, giving in to political bias-led selective coverage and skewed narratives will not only be a disservice to honest journalism but also missing out on doing in India what the Nassar case did to the US.

Balaji Sri Raj contributed to this report.

Author

  • Saikiran Kannan

    Saikiran Kannan is a Data Analytics, Data Privacy, & AI/ML SME, and an Independent Journalist covering conflicts, global affairs, counter terrorism and data stories. He has a Masters in Data Analytics from the Singapore Management University and is a Harvard Business Advisory Council member.

Tags:

Leave a Reply