Team Trump, Team Clinton: Enough Already!


I have stopped reading my Twitter feed. I used to visit it daily. Since this presidential election got underway, Twitter and Facebook just irritate me. Instead of casual updates I see useless 180-character political sentiments.

Not everyone is going to vote the same, I get that. Trash talk goes on, but it can cross a line. I noticed people, who I followed for business content, were posting their biased political viewpoints, tirelessly. It was an endless stream of crass statements. I stopped following the most obnoxious ones. Soon it became a chore. So, I just posted but didn't participate.

I am not alone. A recent Pew Research Center poll that revealed 59% of the respondents find social media stressful when discussing politics. Many did not realize how politically different they were from others.

Earlier this year, MIT Media Lab released some interesting analytics. They noted that, unlike past elections where candidate information came from the campaigns, debates or news organizations, this time information came directly from the candidates. They showed the Top 150 influencers, with Donald Trump leading the list and Hillary Clinton second. The candidates had more influence than the news media and other sources. With the media being accused of bias, voters wanted more direct information.

When measuring social media, you must consider influencers. These are the Tweeters with a large following who inspire retweets. When I read the tweets, I wonder if the person had any influence? Is someone retweeting these statements? Or is it just an echo chamber where those with similar opinions just retweet and like each otherís posts? Instead of influence, it is just a circle of affirmation?

But an echo chamber where voters reside is not a new idea. The public wants engaging content. Facebook has studied this idea and has been criticized for tailoring the Newsfeed to engage rather than inform. One columnist wanted to see what others thought even if she disagreed. She was open to different viewpoints because it influenced her thinking. Her friends must have more constructive posts than some of mine.

If social media is the new avenue, then letís just use follower counts to determine the next president. Or we could measure hashtag popularity #ImWithHer versus #DrainTheSwamp? We would then have to worry that it was rigged. Both candidates are using bots to add followers and handle their streams. Atlantic magazine referred to this election as a bot war. The bots inflate numbers but they are not illegal. They just do some of the heavy lifting.

Comments on blog posts are another form of social media. I prefer comments more than blog content. One trend I noticed is those who are silent supporters generally are for Trump. In a recent Washington Post article, they noted that Trump performed better in online polling than when humans asked the questions. These voters claim the other side is so fervent with backlash they no longer share their ideas. But isnít that the idea of social media?

I wonder how that trend plays into the forecast models? Nate Silver admitted his team was completely blindsided by the Trump primary victory. Maybe we cannot predict these things with numbers and data because we cannot remove the modeler's bias. But what other method do we use? An American University professor suggested there are 13 keys that determine the winner that has been accurate since 1984. He says Trump will win. One of my friends suggested that the polls themselves drove the results and she suggested a more accurate method -- betting odds. Let the voters put their money where their mouths are.

Did you stop following anyone? Did people stop following you? Are you avoiding or embracing social media? Which forecasting method do you trust?

Tricia Aanderud, Senior SAS Consultant for Zencos

Tricia Aanderud, Director, Data Visualization Practice at Zencos Consulting, provides SAS Consulting services to organizations that need assistance understanding how to transform their data into meaningful reports and dashboards. She has co-authored three books with her most recent title "Introduction to SAS Visual Analytics". She regularly shares data visualization tips and SAS knowledge through her BI Notes blog (http://www.bi-notes.com). Tricia has a background in technical writing, process engineering, and customer service. She has been an enthusiastic SAS user since 2002 and has presented papers at the SAS Global Forum and other industry conferences.

Predicting Traffic in the Sun's Shadow

How do you predict the popularity of a one-time event? A look at different ways to forecast road traffic along the path of the eclipse.

Where Sharks Attack in Florida

As Shark Week approaches, we've got some data visualizations to help you determine where it's safe to get in the water and where you will need a bigger boat.


Re: I've Dumped A Trump Supporter
  • 1/5/2017 6:26:55 PM
NO RATINGS

@Jim:

> But why spend five minutes trying to be right when you have data that sounds good to you and your political/social mindset?

I think that's how a lot of data science gets practiced in enterprises today.  ;)

Re: I've Dumped A Trump Supporter
  • 1/5/2017 5:37:16 PM
NO RATINGS

..

Seems to me one of the core themes of Tricia's initial blog post was the political polarization of the USA. On Jan. 17th and 18th, PBS's Frontline series is going to feature a two-part special report on this very issue, titled Divided States of America.

I dunno if it will deal with the social media issue, or how the polling led to such wrong predictions, but it's probably a relevant informational input into the process of trying to fix the glitches.

..

Re: Election-time SomeEcards
  • 1/5/2017 11:13:15 AM
NO RATINGS

True, especially with bots now crawling Twitter - I am struggling to imagine if Twitter has some dark days ahead. But people do achieve advocation on social media. Pwe institute notes as much as 20% or people surveyed do change their minds.

Re: Election-time SomeEcards
  • 12/28/2016 9:33:55 AM
NO RATINGS

It might also be argued that tweets in general are really a sort of false news. Politicians are not noted for universally telling the truth or what they really believe. If we could " use follower counts to determine the next president," that might be a way of measuring popularity or even notoriety, but it wouldn't be a way to find the truth of the matter of if those social communications are actually useful for getting at the truth.

Fake news has become a serious threat
  • 12/27/2016 10:14:48 PM
NO RATINGS

..

Jim writes


Fake news in a time of crisis is nothing new. The challenge is in understanding how quickly it can spread and getting people to validate those rumors/news reports. I think of it today as the Facebook data problem, those statistics that live on and on despite being easily disproven.


 

A2 has hosted a number of discussions about the validity of data, the reliability of information and news, and the early signs of the emergence of the "fake news" issue.

Now, especially during and now in the aftermath of this year's catastrophic presidential election, Fake News has ascended as one of the really hot and persistent issues engulifing not just politics, not just journalism, but the contextual societal perception of reality and what's true and what's false ... and whether anyone can really tell the difference anymore. I wonder how this will pan out in situations like courtroom litligation, where the validity of evidence is frequently a central issue. Should be interesting ...

And in case you've missed the news (presumably not fake), the far right has seized on the buzzword of Fake News to brand virtually all mainstream media (plus any other published news they dispute) with the term. On Christmas Day, the New York Times published a report on this: Wielding Claims of 'Fake News,' Conservatives Take Aim at Mainstream Media.

And these same forces aren't stopping there – they've turned their sights on fact-checking efforts, as well, such as the Snopes.com site: For Fact-Checking Website Snopes, a Bigger Role Brings More Attacks.

Take a few moments to contemplate the implications for scientific research, data analysis, and similar pursuits in a country where there is wide general disagreement on what's real, what's fictitious, and whether there's anything that can be described as true or false or right or wrong. Sounds a bit like a science-fiction nightmare, but that may be where the USA is headed ...

Happy New Year ...

..

Re: I've Dumped A Trump Supporter
  • 12/27/2016 7:54:39 PM
NO RATINGS

Will have to check that book out.  Thanks for sharing it.

Re: Election-time SomeEcards
  • 12/1/2016 8:53:14 AM
NO RATINGS

@Pierre. Fake news in a time of crisis is nothing new. The challenge is in understanding how quickly it can spread and getting people to validate those rumors/news reports. I think of it today as the Facebook data problem, those statistics that live on and on despite being easily disproven.

How old is fake news. After the battles of Lexington and Concord, word spread throughout the colonies that the British soldiers were "scalping" the Minutemen. That atrocity was easily disproved by witnesses who said the couple incidents of so-called scalping were actually cases where there was hand to hand combat and some Americans were killed by soldiers swinging empty muskets like clubs. Of course, the myth lived on, perpetuated by the "tavern talk" that was the Facebook of the day, and some Americans continued to take revenge on the British soldiers in the years that followed.

That could have been avoided if they just had Google.

Re: Election-time SomeEcards
  • 11/30/2016 9:34:15 PM
NO RATINGS

We're going to see a lot of dialogue about cyberwarfarem especially with fake news being news.  How can a democracy operate when its citizens are sending fake concepts? Not well.

Re: I've Dumped A Trump Supporter
  • 11/30/2016 9:31:33 PM
NO RATINGS

I think that is coming, though a challenge is sifting real sentiment from bot.  There was a lot of debate about how many real follower profiles Trump had during the election.  Some were bots that could even respond. I recall when just ahve any sentiment on Twitter was questioned.  Sifting through real response versus bots that have been able to respond as aperson is now the barrier.

Re: I've Dumped A Trump Supporter
  • 11/30/2016 9:27:49 PM
NO RATINGS

But there would really need to be dancing - a lot of dancing. ;-)

Page 1 / 9   >   >>
INFORMATION RESOURCES
ANALYTICS IN ACTION
CARTERTOONS
VIEW ALL +
QUICK POLL
VIEW ALL +