Extreme Honesty a social media experiment

As children, we’re constantly tasked with telling the truth, yet at the same time we immediately begin bombarding our children with conflicting statements: Santa, Tooth Fairy, Easter Bunny, etc. I set out in the fall of 2009 to test the range of emotional responses and aggregate effect on social media accounts of being honest. I set the stage months earlier for the experiment by starting “no filter Friday” an experiment to see whether people could grasp true honesty in a business endeavor. I also piqued the interest of quite a few new followers by conducting a smaller scale experiment. Both went perfectly according to plan.

In midsummer I began planning the biggest experiment yet, an experiment in which I lost the ability to have “balanced rational discourse”. Or to put it another way, I became as intolerant as many of the “popular” twitter accounts I saw. I didn’t lie, I merely took the extreme view of whatever my actual opinions were. Which for those who know me, I’m already unabashedly liberal I just took it to the “if you don’t care about people more than anything else you’re evil” level. That was only a part of the experiment.

The other harder part was to always tell the truth. I did have a major setback in feeling confident about the experiment when, due to a combination of multiple factors, my marriage ended. While it hurt immensely, I eventually decided to resume work on the experiment. I hinted at the experiment in the weeks before it started and discussed it with a couple of my close friends before deciding that come hell or high water on Dec 21st, 2009 I would begin. My twitter account was now ground zero for a whole new type of experiment (years and years ago I lived as a homeless person in Portland, Oregon for 6 weeks because I wanted to see what the actual effects on my own personality would be).

A few people including Pete Grillo, Steven Mcdade and Thomas Brenneke came perilously close to making me ‘out’ the experiment throughout 2010. My one rule to maintain telling the truth while taking an extremist tack was the same as used by Project Alpha: If asked why I was required to tell the person asking. Thanks to a few of you for keeping it secret by the way!

Due to a rather nasty server issue I lost all of the data I’d been aggregating of my own tweets, interactions and follow/unfollow counts. While still a massive setback the raw data numbers weren’t the main focus of this experiment. Detecting the initial problem  (an outdated script incompatible with a recent update to the server) was difficult due the script failing silently. When all was said and done I had the first 3 months of data and the last 5 weeks, the interim data was simply gone. I was faced with a tough choice: Throw away a lot of work that did alienate friends and coworkers or do an accelerated version of the experiment in a single 24 hour period. Since I believe that knowledge is so important I took it upon myself to regain the lost data. I did so the second Saturday in January and took it to a much higher level than I had before. Ratcheting up the rhetoric without crossing my own personal line of advocating or threatening harm to anyone. Given the heated nature of the topics a few friends reacted hurtfully (and with logical reasoning too).

Unsurprisingly the end of the year conclusions came out fairly staid:

• People under the age of 25 are vastly more likely to positively respond to reactionary statements.
• People identifying as monotheists are more likely to unfollow when their ideologies are questioned or insulted than polytheists or atheists.
• People who use their faces as avatars are less likely to deviate from a moderate tone in their tweet stream.
• The amount of unengaged accounts went up at approximately the same ratio as my daily number of tweets. While many of these accounts aren’t necessarily spam it does indicate that neophyte users frequently just follow anyone in the stream who says things that they relate to. Nothing out of the ordinary there really.
• The number of spam accounts did spike rather heavily (up 23% the first week no less) as I began using more hashtags and keyworded phrases.
• One fact I’m quite proud of is that my fellow Portlanders were the group least likely to unfollow me based on my statements. When I looked at the data closer the numbers of “loyal” followers were split fairly evenly between liberal and conservative users. Which led me to conclude that Portland is more about speaking up than it is about partisanship. Not everyone in Portland may agree but they won’t necessarily unfollow you for it.
• At the 6 month mark I ramped up to following a large number of adult performers, strippers and cam girls throughout the world, again surprisingly Portland accounts took offense at a much lower rate than other cities.
• At the 9 month mark I began cycling on a 48 hours rude and 24 hours polite rate which resulted in very little differences to the statistics over time but did cause some upheaval during the switch from one state to the other.

The few really interesting realizations occurred in the way many of my long term friends and some quiet acquaintances changed the way they saw me over that year. Especially interesting was the group of people who publicly said one thing but privately said another. It speaks volumes about how even though we’re told to tell the truth as children, between the age of being a child and being a fully fledged grownup we lose those abilities to always be honest.
Of my personal favorite responses and results were:

• An active social media using tech executive from California regularly sent me a DM whenever I took certain topics to task thanking me for “saying what he would get fired for”.
• A VP of Marketing from a NYC based company sent me an email that read in part “Some of what you say is obviously calculated to get a response but nicely doesn’t come off as blind trolling. It would be nice if honesty of the level you approach was something marketers could capitalize on”.
• A former student from my drumline teaching days called me to ask if I was really that angry at the world and if there was anything they could do to help.

After going through the data and realizing that a lot of what our parents taught us is true, that how you speak affects how well you’re listened to I realize that the greatest lesson wasn’t the specifics of the data. My experiment was never about collating the many into a grand unified theory of sociology, it was about how the circle of friends and acquaintances I have change and adapt over time. Everyday our personalities go through varying degrees of flux. It’s a natural part of being human, what we must all learn to do is be honest to ourselves and to everyone else without remorse. It’s not a real decision if you’re not invested in it.

While I’ve got much of the data aggregated now I’m not able to share the full report at this time. When you blindly involve people in an experiment you don’t have a right to arbitrarily decide how to use their public identity. I’ll try to figure out a way to sanitize it enough to not force many people to have their information shared as part of this. I should add that I also am under an NDA which prevents me from releasing anymore data at this time. I will say thanks to the very cool company that liked the idea and paid me for the researchas well as provided me excellent resources for getting the missing data restored as much as possible. Thanks guys!

I feel enriched and also deeply humbled by the gamut of emotional responses to my words, my deeds and especially the number of new friends I made during the experiment.
To those of you who helped voluntarily or as subjects I say thank you very much.