Friday, January 23, 2015

How to improve team intelligence


Slashdot posted an intriguing article about teamwork and why some teams outperform others:

http://science.slashdot.org/story/15/01/19/0249200/why-some-teams-are-smarter-than-others

The original study is here:

http://www.chabris.com/Woolley2010a.pdf

The basic findings is that intelligence of any individual member doesn't make a team better nearly as much as three other factors:

1. The average social sensitivity of group members, i.e. how well group members could read the feelings of other members.

2. How often each member spoke. Groups where one or two members dominated were not as good as others where speaking turns were evenly distributed.

3. The proportion of females in the group.

This science, assuming it's valid, proves a lot of things that the Navy gets right, and should guide us on what we need to change to improve in the future.

The Navy gets a lot of teamwork ideas right.  The first one that comes to mind is planning teams.  I'm a recent graduate of the Maritime Staff Operators Course (MSOC).  I worked with other Navy Officers from across the spectrum, from JAGs and HR officers to SWOs, Submariners and aviators, and even the occaisonal EOD guy.  Working together was challenging, and we didn't always get along, but overall our class was successful in planning and executing a campaign.  I found the best plans were made when we were given great leeway and broke up into small groups of 3-4 people.  When the groups were bigger, too many people had to compete for a chance to speak.  When we received heavy direction on the desired Course of Action (COA), we made a plan, but it wasn't very creative.

I also feel like the Navy values Sailors opinions, at least in many areas.  I've commonly heard "the best ideas come from your Sailors."  I don't think the Army and Air Force say quite the same thing, but it's common knowledge in the Navy that the junior petty officer or lieutenant will come up with some bright idea that will often solve the problem.  The Navy's culture of pushing decisions down to Chiefs, Junior Officers and Enlisted is a good thing, as it lets good ideas work, at least locally.

I feel the Navy struggles most with rank when it comes to team collaboration.  Too many people use rank to squash ideas at the outset.  I've personally watched as junior personnel bring up good ways to change, only to have an Executive Officer or Commanding Officer (or Chief, or Department Head, or Division Officer) squash it outright.  In those situations, it was no surprise that after about 4 months, only the XO and CO ever came up with anything new.  In one case, the CO openly complained that no one brought him good ideas.

(Courtesy of Dilbert.com. http://dilbert.com/strip/2008-04-27)

Rank gives us the means to make things happen.  Used correctly, rank is an extremely powerful way of creating change quickly when we need to make something happen fast.  Think about a fire on a ship with a fire team and a man in charge.  There isn't time to debate, so the man in charge makes it happen by leading a team right there, on the spot.  It's efficient and extremely effective.  I've seen COs end years of problems by a simple stroke of a pen because they knew what sort of power they had.  I've been fortunate enough to be given the leeway to use my rank to make positive change at many commands now.  Rank is a good thing.

The way forward with this study is two-fold.  First, we should be training people to read emotions better.  I wouldn't call it sensitivity training, mainly because that is fluffy and will be taken the wrong way.  Being able to read people's emotions in their faces though is obviously a critical skill to successfully running a team.  Until the Navy makes some sort of training, you can start by reading some books on the subject.  The one that immediately comes to mind is Emotions Revealed: Recognizing Faces and Feelings to Improve Communication and Emotional Life by Paul Eckman.

Second, we need to improve our collaboration tools to get more participation from people.  We need to recognize that not everyone contributes well in a group setting, especially if it is dominated by extroverted people.  As someone that falls in between the Extraversion and Introversion scales on the Myers & Briggs, I know that face to face groups work great if you're an extrovert, but sometimes the quiet guy or gal in the room has the best idea and just won't be heard.  The Navy openly values extroversion, but if 50% of people are introverted, then we set ourselves up for failure by only listening to half of the good ideas.  Making changes to our group formats (e.g. ensuring everyone speaks) or using other formats (e.g. online forums) will probably get us better results by getting input from everyone.

I'm not thoroughly convinced by the increase in females for better collective intelligence.  Even the report says "However, this result appears to be largely mediated by social sensitivity (Sobel, z=1.93,P=0.03), because (consistent with previous research) women in our sample scored better on the social sensitivity measure than men [t(441) = 3.42, P=0.001]."  Better training would eliminate any gap that exists between the genders.  Another thought comes from from a Slashdot.org comments:

"I'm starting to wonder if all you jokers read the same paper FTFA as I did, and not just the article. The paper points points out in no uncertain terms that the inverse correlation between group performance and participation dwarfs the (almost insignificant by comparison) correlation between number of women and group performance.

Is there a correlation between number of females and group performance? Yes, but it's only marginally stronger than the correlation between the highest-IQ of the group and group performance. The inverse correlation (-0.41,0.001) between group participation and group performance is a good deal larger AND highly statistically significant compared to the extremely weak correlation between number of women and group performance (0.23,0.007) which is merely statistically significant. And really, a p-value of 0.007 when they only tested 600 odd people (not 600 odd groups)?"


Although I'm not a statistician, it again seems to point towards something that can be easily overcome by training.

We can make teams better in the future, but we're going to have to challenge the notion that the loudest person is the most correct and that being cognizant of others emotions is somehow a bad thing.

No comments:

Post a Comment