Blake Dowling: Elections, psychometrics and social media’s dark side

psychometrics

The last presidential election was unlike any in my lifetime.

All predictions and polls wrong. FBI investigation right before Election Day. Crude language about women right before the election. A rogue candidate vs a political elitist. New revelations of social media ads bought by the Russians to influence voters.

The experts got it wrong, the entire time, as detailed in the Showtime documentary “Trumped.”

There are lots of obvious reasons why this happened, as well as not-so-obvious ones.

Then there is Alexander Nix, Cambridge Analytica and Psychometrics.

Psychometrics is the measurement of psychological traits and the “Big Five”: Openness, Conscientiousness, Extroversion, Agreeableness and Neuroticism (OCEAN).

Pay attention readers, because if you do not know this tale, it goes far down the rabbit hole, then blows all the rabbits up.

So, the hard part about developing models for the OCEAN traits was info; gathering actual data.

How do you do that during the 70s? 80s?

I will tell you how they do it now: internet and social media.

It makes the data collection a breeze and predicting behavior a real science. It would appear that a guy named Michal Kosinski was the forebear of this type of digital science; he developed predictive models of behavior analyzing likes on Facebook. He could predict your skin-color, sexual orientation, and political affiliation based on what you liked and did not like and of course factoring in other data points, age, gender, etc.

Basically, our interactions with social media are equal to a never-ending psychological questionnaire that allows those with the tools to analyze this data to know us better than our friends know us.

So, Kosinki had his methods down and he gets approached by a guy named Kogan wanting to use this method for a company he is allegedly representing named Strategic Communications Laboratory (SCL).

Kosinski is starting to figure out the power of his work and when he looks up what SCL does.

He then has a Fred Sandford moment: “This is the Big One, Elizabeth; I’m coming to join ya,” etc.

SCL focuses on psychological modeling; one of its focuses is influencing elections. Say what?

Allegedly, this company spun off a new U.S.-based company, Cambridge Analytica.

(Also of note, it is rumored that Kogan changed his name to Dr. Spectre and moved to Singapore.)

I think this column is going to have me end up hiding out in some foreign embassy. It is like a chess game with glow in the dark gummy bear pieces — very interesting and tasty.

Traditional political marketing efforts target people using demographics, targeting an entire subset, like females, for instance. This is not that.

CEO of Cambridge Analytica Alexander Nix speaks at the 2016 Concordia Summit – Day 1 at Grand Hyatt New York on September 19, 2016, in New York City. (Photo by Bryan Bedder/Getty Images for Concordia Summit)

Psychometrics can target white women who like Wu-Tang Clan, drink Jack Daniels, live in LA (lower Alabama), love gladiator movies and reruns of Hee-Haw. (Who doesn’t love those?)

This is very specific, and, for all intents and purposes, very effective.

How do you get to people? They show targeted/paid “dark posts” on social media and try to get people to the polls or keep them away from the polls.

For instance, think of a Miami neighborhood called Little Haiti, dark posts are bought and run to users in this area that focus on the failure of the Clinton Foundation to help with Haiti relief after the earthquakes. Or in an African-American neighborhood with targeting paid ads where Hillary in a comment refers to black males as “predators.” These specific people only see these dark posts.

There are targeted posts for everyone, either driving people to cast a ballot or to keep them home.

What did the Trump campaign pay these people? $100,000 in July. $250,000 in August. $5,000,000 in September.

What did they have before CA, some troll named Brad who was paid $1,500 to build the Trump for President website (sorry, Brad; you stink, man).

OK, so door-to-door Trump supporters could knock on a door, knowing what the people inside believe and give them a quick message that they already know they can relate to, and yes, it’s available on an easy to use app.

They did this concentrated effort in only 17 states — states they absolutely needed.

Door-to-door messaging to a receptive audience, one-on-one targeted posts that would be most likely received positively. This sounds like the political version of the New England Patriots.

Guess who is on the Board of Cambridge Analytica, Steve Bannon.

So, is this all BS? Maybe.

Maybe people just liked the crazy dude with sweet hair. Who knows?

More fake news craziness? It is said CA was also a big player in the Brexit campaign, too.

Maybe the sky is red, and maybe Al Gore was right all along, it is hot as hell for springtime.

Anyway, I don’t know much but I know this is not the last time we will hear about this subject.

Keep on rocking people. And long live those women who love the Wu-Tang Clan and Jack Daniels.

___

Blake Dowling is CEO of Aegis Business Technologies. He writes for several organizations and can be reached at [email protected].

Blake Dowling

Blake Dowling is CEO of Aegis Business Technologies. His technology columns are published by several organizations. Contact him at [email protected] or at www.aegisbiztech.com



#FlaPol

Florida Politics is a statewide, new media platform covering campaigns, elections, government, policy, and lobbying in Florida. This platform and all of its content are owned by Extensive Enterprises Media.

Publisher: Peter Schorsch @PeterSchorschFL

Contributors & reporters: Phil Ammann, Drew Dixon, Roseanne Dunkelberger, A.G. Gancarski, Anne Geggis, Ryan Nicol, Jacob Ogles, Cole Pepper, Gray Rohrer, Jesse Scheckner, Christine Sexton, Drew Wilson, and Mike Wright.

Email: [email protected]
Twitter: @PeterSchorschFL
Phone: (727) 642-3162
Address: 204 37th Avenue North #182
St. Petersburg, Florida 33704




Sign up for Sunburn


Categories