My Favorite Analytics Story…Is About Cake

28 04 2012

Here’s a Cross post of a piece I wrote on the Capital Analytics blog:

One of my favorite stories about talent analytics has nothing to do with talent or analytics. It’s about cake. Well, it’s not really about cake, it’s about how the mind makes decisions, but we’ll get to that.

Let me introduce to you a wonderful experiment run by professor Baba Shiv from Stanford. Professor Shiv works in the field of neuroeconomics – the study of the mind’s electrical impulses and chemistry when a human makes a decision. In 2008, he ran an experiment to understand one aspect of how people make decisions based on data.

The experiment went like this: Participants sitting in a nondescript room were asked to memorize a set of numbers. Some were asked to memorize two numbers. Some seven numbers. (Let’s call one such participant Steve.) After Steve had the numbers in memory, the researcher asked him to walk down the hall to another room and relay the memorized numbers to another staff member. However, that wasn’t the real experiment at all.

While the Steve was walking down the hall, keeping the numbers in his memory, another staff member would “coincidentally” appear in the hallway holding a tray of dessert. This staff member said to Steve, “Oh, you’re participating in our study? Thank you so much! You know we just ended a big meeting and have all of these leftover snacks. Can I offer you some fruit or a piece of cake?” Professor Shiv then compiled the choices of each participant. Who chose the fruit? Who chose the cake?

The results from this study were surprising. The people who memorized seven numbers, were twice as likely to choose the cake vs. the fruit. The more data a person kept in short term memory, the more likely they were to make a decision based on emotion vs. rational thought.

Why would that happen? It turns out that our short term memories are like hard drives – it’s possible to fill them up. Neuroscientists call this, “cognitive load”. While your frontal lobe is busy memorizing new numbers, a new task requiring a decision may not be routed to the rational part of the brain. Sometimes the emotional centers of the brain make the decision.

Let’s connect the dots with talent analytics. The moment that one of our senior managers reads one of our well-researched, completely revolutionary, information-packed decks on human capital, it’s possible that we are setting that executive up to fail. While they are reading and absorbing the new numbers, they are making decisions . But which part of them is making that decision? The rational side or the emotional side? Particularly with talent decisions, where a manager may be more emotionally tied to a decision, it may be difficult – physically difficult – for a leader to make a decision based on loads of data. While they are absorbing data in the frontal lobe, the decision is being made… by the emotional centers of the brain.

I freely admit that I have bought into the idea of cognitive load, and I’ve changed the way I provide data to others because of it. Here are a few lessons I’ve learned from Professor Shiv:

First, don’t defy the laws of physics by cramming tons of data on a single page. Excel does allow us to use a 6 point font, but please pity your reader and don’t use it.

Second, know what decision you are asking someone to make. What information is needed to make that decision? Keep that data, and strip out the rest. Throw the backup data in an appendix if you must.

Finally, know what story you are trying to tell. Does it flow? Does it have a beginning, middle, and an end?

There’s so much more to this topic, but I don’t want to create cognitive load. If this is an interesting topic to you, listen to this radio piece on the experiment and try out new ways of sharing your data. If it doesn’t work, come and find me in New York. I’ll buy you a cup of coffee. And maybe a slice of cake.





Interview with Kaggle.com posted on iianalytics.com

29 04 2011

One in an occasional series of articles I write for the Institute for Advanced Analytics… this is an interview with the CEO of Kaggle.

http://iianalytics.com/2011/04/interview-with-kaggle-com/





Innovation in the aging workforce, WSJ got this one wrong

2 09 2010

The WSJ Heard on the Street section printed an article today, “Older US Workforce Has an Ugly Wrinkle” sparked a few thoughts for my post this morning.  

  • First, was a statistic. 40% of the US population over 55 is working or looking to work, the highest ratio since JFK was president. Demographically I found this interesting mainly because the ratio of 55+ Americans to the population has increased since the 1960s. The open question to me is, does the 40% go up from here, as baby boomers reinvent what retirement means? Or does it go down as the pain of the recession begins to wane and nest eggs are more confidently restored?
  • Next, “on the positive side, the declining share of manufacturing in the economic mix… Should make it easier for many to continue working”.  Let’s stop right here and google the age of the author writing this piece. Innovation will come to manufacturing too. BMW prooved it in this HBR article a few months ago. where the workforce mix of the future was artificially created in one of the most heavily labor intensive lines, and the team innovated around the effects of an older workforce. (a 7% productivity gain to boot thank you very much).
  • And lastly, a comment on higher savings rates. Yes, the WSJ is probably correct that savings rates will increase (that used to be a problem we bemoaned, remember that?) However, depending on your outlook on health care costs, it’s certainly likely that more income is directed to health care, and less to savings.

Now those are just my thoughts, someone who has not yet made it to the 55+ demographic… Please chime in and put me in my place Boomers!

     




If you were a PBS kid, watched Zoom in the 70s and enjoy Pi

21 07 2010

That’s about the only combination of variables that would likely make you smile at this strange video about Pi.  I happen to possess all 3 of those variables, so I smiled. Thanks to http://technoccult.net/ for the find.





So that’s why recovery.org doesn’t suck….

22 03 2010

I was listening to On the Media’s interview with Edward Tufte (the pre-eminent visual data designer) and was pleasantly surprised to heat that’ he’s been working on recovery.org’s web site.  I don’t know the onVia software, so let’s assume that Edward isn’t the only reason the visuals don’t suck, but that he’s working on it is very cool:





Visualizing HR Data Part 1: Testing Trend Compass

22 08 2009

I just downloaded a trial of Trend Compass, which has created similar bubble charting software that Hans Rosling used on TED.com and similar to a Google gadget mentioned in this blog a couple of weeks ago. This is only test data! But I wanted to try it out. I’ve loaded a sample data set in here to see how we could tell the story of the recession by source of applicant and hires.  (If you happen to work for any of the companies in the bubble, again, not trying to assess anything about you, you’re a sample data point). If this becomes interesting enough, we’ll try this with some real data.

 

I don’t know how much the SW is yet, I’ve emailed them (that can’t be a good sign). However, the Google widget should work as well, maybe better, it just has less documentation.

Update: They emailed me back, $10,000 for the application I’m testing or $50 a pop on their site directly. The only advantage I see so far to it is that you can record a wav on top… but you can do that with any desktop tool, so it’s not hitting a homerun for me yet.








Follow

Get every new post delivered to your Inbox.