Our Blog

It is always fascinating to experience a change of scenery for a while. And I was very lucky that our project funds contained money for 3-months research stays abroad. Therefore, I want to explicitly thank the German Federal Ministry for Education and Research for the funding that made this endeavor possible in the first place.

I used the opportunity to widen my horizon while (hopefully) still doing something useful. Taking the CBIE Winter School on modeling human behavior in ABMs in Tempe, Arizona as a starting point, I stayed on at the Center for Behavior, Institutions, and the Environment for ten more weeks. CBIE is part of Arizona State University, with over 70,000 students one of the largest universities of the US. Its largest campus is located in Tempe which is a suburb of Phoenix.

After interesting discussions and lectures at the winter school, I expanded on that topic with CBIE director Marco Janssen through a new research project: Marco had conducted a set of experiments on human behavior in irrigation systems. He had also developed an ABM that mimicked these experiments in order to test different established behavioral theories. Together, we extended the model and are now going to test how well these behavioral theories can explain the empirical data.

For that, I dived into the nuts and bolts of Approximate Bayesian Computation (ABC) which is a novel method that seems promising for different kinds of model validation. What it does can be simplified to this: If we consider all possible parameter combinations in a model as a giant haystack and take those parameter sets that reproduce the empirical data closest as the notorious needle in that haystack, then ABC describes a way to systematically look for that needle. And by using Bayesian inference, it does not only find the needle, but also indicates the probability of each piece of straw to resemble the needle. Or to put it more scientifically, ABC does not only provide best estimates for parameter values, but also their uncertainty.

In the next months, we are going to apply ABC to Marco’s irrigation model and expect to generate some insights into how human behavior in commons dilemmas can be explained. More on this will follow on this blog…

By the way, if you ever plan to visit Phoenix, go there in February or March. The temperatures are warm and the weather is really nice. But by the time I left, it had already gotten up to 36°C and it was getting warmer every day.