2014 was an amazing year for shaping human behavior


This is simply the 3rd installment of my annual summary of everything I tweeted about shaping behavior over the year. From statistical innovations to physical shop design, a lot of interesting things made it to mainstream news and made 2014 a really interesting year. Although it would have been great if it didn’t, even A/B testing made it into the headlines with Facebook and OkCupid’s public relations disasters.

Although I mostly tweet about online behavior change and conversion optimisation, this year was also surprisingly much about behavior change unspecific to on- or offline. While the CIA manual to creating and spreading rumours is from 1944, it came onto my radar this year through a talk on information security by Rand Waltzman at the KDD conference, and it’s a fascinating read and possibly as useful in 2014 as 1944.

I’m @matseinarsen at Twitter, if you want to follow me for more updates in 2015 or want to recommend an article.

Shaping real world behavior

First of all just a few articles on offline behavior shaping. I really love this perspective, and wish there were more things published about shaping group behaviour. On the other hand, I take it for granted that companies protects insights in this area as highly important trade secrets, so not expecting much except the occasional drip.

Experimentation and statistics

Two really fascinating things came about this year. First is CodeCon — the Conference on Digital Experimentation — which was held for the first time at MIT this august. I can’t state enough what a great conference that was. Check out the link below to the beautiful brain maps that were drawn during the conference.

The second fascinating thing is the noise analysis methodology to identifying cause-and-effect relationships in correlations. The methodology is really straightforward and makes intuitively sense, and even in the case that you can’t make strong statements about causality it will at a minimum help in prioritizing the importance of correlations when going fishing in big data sets.

The Facebook and OkCupid controversy

This story has annoyed me as there’s been a lot of statements about ethics in research I just don’t agree with and basically find to be very muddled thinking. Not to make a statement about the ethics of the Facebook experiment (ethical) or the OkCupid experiments (mostly not ethical), but overall I don’t get why the experimentation itself is at the centre of the ethics debate.

Would the changes Facebook or OkCupid made be more or less ethical if they didn’t have a control condition and measured the impact? I can’t see why that makes any difference. And then you’re just left with a debate on whether or not companies can do unethical things and that is just clearly the case and not a very interesting debate on it’s own.

Either way, here are the most important articles written about it over the year.

Deals and Decisions

The role of self-interest in elite bargaining
Humans frequently act contrary to their self-interest and reject low offers in bargaining games. Some evidence suggests that elites, however, are much more rational and self-interestedm.pnas.org

Online persuasion and conversion optimisation

Social effects

Data Science Cuisine

Ok, I just have three things here this year, one of them isn’t a proper article, and “Data Science Cuisine” is just a term I made up myself. But I’d really like to see this become a thing! Please everyone, let’s make data driven gastronomy happen!

Data-driven business

Putting this together, I realize all the articles I posted on this topic this year are from Harvard Business Review. I guess there’s an endorsement of HBR in that, but it would be nice to see good, academically driven content about business topics from another source too. The two slideshows below on Google and Netflix are quite interesting insights from the insides of important companies, though, so highly recommended.

Design and creativity

Brain hacking

DIY brain hacking. More of this too, please!

The below open platform for EEG measuring has my friend @Eric_Herman involved and the aim is to make an open source and open hardware EEG controlled mouse-pointer. I can’t wait to see this happen.

Psychopathology

The most interesting point to bring up here is how the observation that Twitter usage can be linked to psychosis — should that influence how we look at the ethics of AB testing in social media?

Big summaries

I like this guide so much I wanted to give it special attention.

And finally, for reference since these were on my private blog, here’s my summaries for 2013 and 2012: