Dec 08 2017

In Half a Second

If you have not yet read Daniel Kahneman’s book, Thinking, Fast and Slow, I recommend it. I have discussed its basic principles here many times, and I am reminded of it by a new study that evaluates how we quickly size-up groups of people.

Before we get to the study, here is a quick overview. Kahneman and Tversky did the foundational research into cognitive biases and heuristics – ways in which our thinking is biased or constrained. Kahneman calls this system 1 thinking, or intuitive thinking, which is the fast sort. There is also system 2 thinking, which is slow and analytical.

He admits that these are metaphors, there probably aren’t two distinct biological systems in our brains, but they help us think about the different ways in which we think. Actually, given that our brains are hierarchical, the two-system model may be based in biology to some extent. There are the primitive older parts of our brain that are more system 1 – instinctive, emotional, and fast. Then there is the neocortex – which gives us executive function, and slow deliberative decision-making. I don’t think you can make a clean separation, but it is a useful schematic that is probably more true than not.

In any case, the two systems work together to shape our perceptions and decision-making. The idea is that we evolved rapid-response cognitive systems that makes quick and dirty judgments that are accurate enough and biased in whatever direction favors survival. We can then follow up these quick perceptions with more careful analysis when we have time.

But also, we are fundamentally lazy, which we can less judgmentally characterize as being efficient. We seem to always be looking for ways to minimize our cognitive work. So we rely on system 1 automatic thinking probably more than we should. We also have a tendency to substitute easy problems for more difficult ones.

For example, we have a left number bias. There actually is a reason that items cost $19.99 instead of $20.00. We know rationally that the penny difference is insignificant, but we don’t spend the mental energy to think about it. We have a tendency to substitute a simpler algorithm – how big is the left most digit, for the slightly more involved task of analyzing the entire number. $19.99 feels smaller than $20.00. This is a reasonable short cut, if numbers are random, but the short cut is easily exploited by crafting prices that end in all 9s.

Perception of Groups

With that background, let’s take a look at a new paper published in Social Psychological and Personality Science called, “Threat in the Company of Men: Ensemble Perception and Threat Evaluations of Groups Varying in Sex Ratio.”

Previous research has found that people can very quickly determine what general category another person belongs to – male, female, race, and age. We can also very quickly (less than a second) determine someone’s emotional state – are they happy or angry.

This makes sense from a survival point of view. An angry young adult male is likely to be more of a threat than a happy old woman. We can almost instantly make that determination, and then dedicate more resources to further evaluating the threat represented by the angry young male. Are they armed, are they looking at me, are they alone?

This is also interesting because perception research has shown that sometimes the age, sex, and race of a person is all we really perceive about them. You can see videos demonstrating this online. About half the time, if you switch one person for another during a casual social interaction, the subject won’t notice the change. The probability of detecting the change in person goes up if you change age category, sex, or race.

What this may reflect is that we don’t notice the change if there is no change in perceived threat category.

The new study extends this research to groups. They had subjects quickly view pictures of groups of people and then asked them various things. They asked them to estimate the ratio of men to women, and to evaluate how threatening the group was. The researchers also measured indirect implicit markers of threat perception.

They found that in 500 ms, half a second, subjects were able to accurately perceive the male:female ratio of groups. This ratio also was the main factor in determining their assessment of how threatening the group was. The total number of males was also a factor.

What this suggests is that our system 1 easy problem that our brains use to rapidly assess potential threat of a group is how many men does the group contain in total and what is the ratio of men to women. We can get this information in a flash.

And again, it is easy to make sense of this in terms of survival advantage. Primate war parties generally consist of all males. If you come across a group of 20 males, they are probably up to no good. A mixed group of half females, however, is probably not going to war or hunting, but may just be a traveling social group.

Of course this is not always true – that’s the point. It is just true enough to lead to a quick 500 ms snap impression that will alert us to a potential threat. We can then take a closer look (from a hidden vantage point, perhaps) to more thoroughly evaluate the potential threat.

By coincidence I happen to be watching Godless – an excellent miniseries, by the way. This takes place in the old West and in part follows the exploits of an outlaw band of thirty men. The premise of the show is that the West at this time was a very dangerous place. Any time strangers meet they immediately suspected danger and were extremely cautious, and their assumptions were usually warranted.

In several scenes characters come across the band of 30 outlaws, and the show does an excellent job of creating the impression of how intimidating and threatening such a group would be. I definitely noticed watching this show how as a viewer you start to size up groups of people to determine if the characters you are following are being threatened, and you can see how any large group of men is immediately suspect. Even when they are the good guys, they are menacing until proven otherwise. Likewise the presence of women in the group is reassuring.

I also notice that the presence of young children in large groups is also very reassuring, although this was not examined in the current study. This research could also be extended to include the age of the people in the groups, and also their racial makeup.

Beyond just being fascinating, research like this elucidates how our brains function. This knowledge can then be turned inward – to tweak the relationship between system 1 and system 2 thinking to maximal utility. Ideally we would extract the advantage of making rapid assessments, but know how and when to back them up with analytical thinking, and avoid succumbing to lazy thinking.

192 responses so far