How angry was CNN after Donald Trump won the 2016 presidential election?

January 1, 2018//1,702 words

There are a lot of questions about truth and fairness that we have regarding major news media stations and the articles they post, especially during the past couple of years.

Some have turned into full-fledged wars (Trump vs CNN). Others just cause fear and doubt – how truthful is this article that I’m reading online? This other one is making some pretty big statements – is it real, or overly-dramatic in an attempt to force an agenda? This one website says that President Trump’s presidency has been a success so far, this other one says that it’s been a failure. Which one is telling the truth?

Well, I can’t speak for the character of journalists, if they truly produce unbiased work, or their ethics. I can say that I have been the victim of false-reporting personally, during two separate stories in my lifetime to this point, so there’s that. But there are more than two reporters out there, so for the ones I don’t know, let’s take a data-driven approach.

I recently launched a project online called “Before You Post That” that uses the power of IBM’s Watson to analyze text and recognize different tones of voice (e.g. confidence, anger). I’ve encouraged others to use this tool to analyze important emails before sending them. For example, before emailing your cover letter to a hiring manager, make sure it comes across as confident, not tentative.

But I thought I’d take a different approach this time – what can IBM’s Watson tell me about the tone of voice of a few major media outlets during and around major news stories (e.g. the 2016 presidential election)?

So, that data is contained in this brief post. I looked at posts from November 7, 2016 (the day before the presidential election) and November 9, 2016 (the day after). And I looked at some of the nation’s largest and most popular political sites – CNN, Fox News, NBC News, RealClearPolitics, FiveThirtyEight, and Politico. I don’t know your preexisting perceptions of these sites, and I only have a few (I only follow politics enough to be somewhat aware), but let’s just see if the data aligns with or differs from these perceptions. And then you can do with it what you will.

Process: I performed Google News searches for the specific dates, specific domains and included the keyword “politics”.

Research Data

Fox News

fox-news-tone-of-voice-2016-presidential-election

Highlights:

CNN

cnn-tone-of-voice-2016-presidential-election

Highlights:

NBC News

nbc-news-tone-of-voice-2016-presidential-election

Highlights:

RealClearPolitics

realclearpolitics-tone-of-voice-2016-presidential-election

Highlights:

FiveThirtyEight

fivethirtyeight-tone-of-voice-2016-presidential-election

Highlights:

Politico

Highlights:

Recap

The summaries are sort of deceiving. It makes sense that large sites like Fox News and CNN produced more articles on the two days in question than smaller sites like FiveThirtyEight and RealClearPolitics. Because of that, individual articles by the smaller sites held more weight in the summaries.

Additionally, posts that featured quotes or input from other individuals likely realized a larger range of extreme tones. But the company is still responsible for what it posts on its site, regardless.

I found it far more telling to look at the individual articles.

Fox News was most joyful when speaking negatively about Hillary Clinton. And they were most sad when, before the election, they reported that Clinton’s lead in the polls was growing.

NBC News was most angry when discussing Trump and receiving congratulations from Vladimir Putin, and when reporting that Trump received the highest amount of votes from working-class white people.

Follow me on Facebook and Instagram to let me know what you think. I’m interested in doing more just like it, so stay tuned …

Also, if you’re interested in doing a study like this, you can access a special section of the “Before You Post That” tool for multiple URLs by going here.