How it Works

Balanced Media is a multi-faceted news rating application, which includes a browser extension, whose goal is to find articles that include multiple perspectives. When a user visits an article on a news website, a popup shows up. This popup prompts the user to rate whether the article includes their perspective as well as how biased the article is. Using some data analytics, we can determine two things:
What the political bias of the user is.
Whether the article contains perspectives from across the political spectrum (or on the contrary, if the article is one-sided).

Using all the data collected from the extension and user, the extension recommends articles that take the most balanced approach - articles that include perspectives from across the political spectrum. These articles help the user to consume a balanced media diet. Furthermore, it rewards news outlets for being more inclusive in their approach to promote understanding rather than peddling outrage. It’s similar to food in some ways - food packaging is required to have nutritional labels. Content should also have nutritional labels. Everyone knows that if you eat only one category of food it’s unhealthy. Balanced Media is that nutrition label. It’s unhealthy to fill your thoughts with only one perspective. Balanced Media’s feed helps people get the media nutrition they need, and break free from the echo chamber of ignorance. My hope is that the influence of Balanced Media will help us as a country to start dialogue, come to consensus, live more peacefully with each other, and all get a little more of what we want.

AllSides is a great tool which helps people to learn the political bias of what they’re reading. However, AllSides doesn’t solve the root problem of helping people to learn about other perspectives - they just help people to see their own media blind spots. There is currently no system out there that attempts to rate articles based on whether multiple perspectives are contained. Balanced Media does this through crowdsourced data. With this data, we can find the most well-rounded articles that include a spectrum of perspectives. Once people start recognizing other’s beliefs and values, it becomes much easier to solve problems through dialogue and consensus.

You may agree that nowadays, conversations with people of different views that turn political never end well. It is our belief that this stems from a deep divide that has been widened through the echo chamber the internet has created - the click through economics that drive media to push out outrage fueled content. The best way to solve this problem is by helping people to recognize the beliefs and values of others.

So many problems stem from division, and so many solutions are hindered by division. After seeing the capital riot, it led us to wonder how people became so devoted to something that they were willing to storm one of the greatest symbols of our Republic. It wasn’t just a fluke, this was just an effect of indoctrination and polarization from an echo chamber. This came just months after a summer of violent protests of the BLM movement and clashes with counter protesters. Clearly, there are problems with elections and racial inequality. But, they can be best solved through understanding and dialogue. Once we dug out the root problem, the next question came up - what’s the best way to promote understanding? Hence, the idea for Balanced Media was born.

Balanced Media grades articles in 2 ways.

The first way is the bias of the article (e.g. left leaning, right leaning). This is on a scale of -2 to 2, where -2 is very left leaning, 0 is neutral, and 2 is very right leaning. When users input what they think the bias is, it comes in as a -2, -1, 0, 1, or 2. However, in the Balanced Media database, the bias can be a decimal number (e.g. 1.5). Instead of simply averaging all of the users ratings to determine the bias of an article, Balanced Media clumps the data based on users’ self-identified political leaning, and then takes the average of that. The process looks like this:

The second way Balanced Media rates articles is through its containment of perspective. The question asked is “This article was written in a nuanced way that includes my perspective on the issue.” The scale is strongly disagree: 0 to Strongly agree: 4. For each article, 3 averages are taken, one for left-leaning users, one for neutral users, and one for right-leaning users.

For the article shown, users who self-identified themselves as left-leaning believed the article strongly included their perspective, with the average being 3.5. Users who self-identified as neutral believed the article included their perspective, with the average being 3, and users who self-identified as right leaning believed the article slightly included their perspective, with an average of 2.5. This article is well balanced, with all sides believing the article included, to an extent, their perspective.

Learn more about Media Bias: https://www.mccneb.edu/Current-Students/Current-Students-Resources/Library/Web-Sites/Media-Bias.