1. Skip to content
  2. Skip to main menu
  3. Skip to more DW sites

Can we predict human decisions?

January 12, 2015

We ask Dr. Thorsten Pachur from the Max Planck Institute for Human Development to what extent we can predict how people - and burglars, in particular - make decisions.

https://p.dw.com/p/1EI21

DW: Thorsten Pachur. So starting off with software? What's your opinion of this software. Can it really help fight break-ins?

Thorsten Pachur: Well, apparently, it does. Concerning the success of that tool, it really was helpful in decreasing crime rate to a certain extent. Based on that empirical data, it works. And actually it's not very surprising that it works because the guys have done a job extracting from past behaviour, using all the information, the data that is out there to extract regularities in the behaviour of burglars.

DW: Now you study how people make decisions and judge risks. Let's talk about criminal behaviour and burglars specifically. How does a burglar make a decision to rob a house, for example?

One of the studies that has specifically looked at decisions by burglars when they make the decision about where to break in is that they consider relatively few pieces of information about a potential property where they could break in. These characteristics specifically concern: what can be gained from a particular house? Whether there are any signs that the people living there are rich. And, on the other hand, they also consider the risk of detection: whether there are any obvious alarm systems are installed, or whether there are any bushes to hide in. So they really try to weigh the benefits and the risks when making these decisions.

Now that sounds rather superficial when you look at it from a moral point of view. What about consequence? Do criminals have another sense of consequence regarding their actions?

That's one hypothesis in the literature at least. And there have been a couple of studies that have specifically looked at the weighing of risks by people who end up in prison and a control group and indeed some research has shown that people who have done criminal acts have different risk perception, so there seems to be some difference between these people and people who do not act criminally.

Now you've conducted extensive research with custom officials at a couple of airports in Switzerland. Now tell us more about that work and how customs officials target certain individuals.

What we have shown in this research is that customs officers use very simple strategies when they're making a decision about whom to search at an airport and they also have to do that as they don't have a lot of time to consider a lot of information about potential passengers who should be searched. And interestingly our research also has shown that people who do not have the same expertise as these customs officers they use much more elaborate strategies. So this indicates that in order to use a very simple and efficient strategy you need to have knowledge.

Alright, well, we're going end on that. Knowledge is alway a good thing. Thorsten thank you very much for this discussion today. It's been very interesting.

(Interview: Meggin Leigh)