Thursday, January 13, 2011

Intimidating Awesomeness

A few days back in the office, at about 9pm, the conversation goes something like this:

A: Hey, try to solve this. My friend was asked this in an interview. You have a stream of words coming in. How will you check, in constant time, if you have already received a particular word or an anagram of it?

Now, if you are wondering what’s with the title of this post, I suggest you stop here and try to solve this problem before reading further.

(Don’t look below.)

B: Hmm. (pause) Well. You could maintain a hash table. Hmm. So that reduces your question to what the hash function should be.
C: We could assign a numeric value to each character and add it up.
D: Yeah, but we’ll get a collision when letters repeat in a word.
C: (pause) Wait a minute. Why only add? Why not some other operation?
B: Like Multiply?
D: Prime numbers.
C: Yep, map each letter to one of the first 26 prime numbers and multiply them.
D: That should always give a unique number.
A: Correct answer!
E: (walks by) Dinner?

All in under a few minutes.

If this were a real tech interview, a successful candidate for a high-end product development job would be expected to solve (or come reasonably close to solving) this in around 20 minutes. Ok, granted we ganged up on the poor problem. But I can’t give you a real work example here, so this is all we have. And, more importantly, that fact that we ganged up on the problem does not explain everything. Yes, the folks I work with are smart, some of the best that money can hire. But then, from experience, I know that most other groups of equally smart people can not pull off something like this. I’d say that what you just saw there was more than the sum of our individual abilities.

Collective Intelligence, the measure of a small group’s problem solving ability, depends on more than the intellectual abilities of individual members. A recent study by Woolley et al. published in Science and appropriately titled “Evidence for a Collective Intelligence Factor in the Performance of Human Groups” shows there is such a thing as a measurable collective intelligence. That it is not strongly correlated with the average or maximum individual intelligence of group members or the traditional soft factors like group cohesion, motivation, and satisfaction. And that it is correlated with the average social sensitivity of group members, the equality in distribution of conversational turn-taking, and the proportion of females in the group.

Social sensitivity was measured using the ‘reading the mind in the eyes’ test, where you are shown photos of the eye region and asked to choose which of two given words best describes what the person in the photo is thinking or feeling. It essentially tests your ability tune in to the other person’s mental state in order to understand or predict their behavior. What is otherwise called empathy, ‘social intelligence’ or ‘theory of mind’. Needless to say, groups of people with higher sores on this test did better when working as a team.

The proportion of females in the group was another factor. While Woolley et al. suggest that this could simply be because women in general score better on the social sensitivity scale; I’ll make a wild assed guess (WAG) and speculate that it has something to do with our neurobiological response to the opposite sex – something perhaps to do with oxytocin in the brain – that causes a gender diverse group to be more socially sensitive.

And not surprisingly, groups where a few people dominated the conversation were less collectively intelligent than those with a more equal distribution of conversational turn-taking. If you look at our sample above, the conversational turn-taking is striking with folks completing each other’s sentences towards the end. Sort of like what Hasson et al. describe as neural coupling, where communication is a single act resulting from brain wave synchronization across participants.

A smart team that scores high on social sensitivity will beat a team of individualistic superstars hands down any day. So, what can you do if your team is not exactly overflowing with social sensitivity? Fortunately, there are workarounds, like the improv technique for brainstorming that Scott Berkun describes in his classic project management reference Making Things Happen. Based on the principles of improvisational theater, this technique forces equality in distribution of conversational turn-taking in a constructive way and thus compensates for lack of natural social sensitivity. In fact, I’ll take another WAG and say that practicing improv could possibly improve social sensitivity skills.