5. Conclusion and discussion
The aim of this study is to better understand the emergence of conventions in situations where individuals have different preferences. Previous experimental research has shown that agreeing on a convention is more difficult in situations where actors have different preferences (Bojanowski and Buskens, 2011; Helbing et al., 2014; Hernández et al., 2013; Sekiyama, 2014). However, most of this research has assumed that people are not able to choose different conventions when interaction with different partners. Societies are changing rapidly and people tend to invest more in flexibility, for example by learning how to work with different software programs. This flexibility might be especially useful when people have to work together with multiple partners that have different preferences. We analyzed decisions of actors in networks on, for example, which operating system they want to use for designing a new program with their different neighbors. We studied heterogeneous populations in which some subjects preferred coordination on one behavior and other subjects preferred coordination on another behavior. As experimental manipulations, we varied the distribution of preferences within the networks and whether subjects were able to differentiate behavior when interacting with different others or whether they could invest in such a flexibility in behavior.