Journal on Policy and Complex Systems
Figure 5 . J1 = J2 = J3 = 0.6 , K12 = - 0.2 , K21 = 0.2 , K13 = 0 , K31 = 0.2 , K23 = 0 , K32 = 0.2 .
Figure 6 . Attractor . J1 = J2 = J3 = 0.6 , K12 = - 0.2 , K21 = 0.2 , K13 = 0 , K31 = 0.2 , K23 = 0 , K32 = 0.2 .
How can the model be populated with real data and used to study a specific conflict ? We have already used poll data to assess initial intra-group attitudes and linkages and inter-group linkages in the two-group model ( Diep et al ., 2017 ) that explored 2016 conflicts surrounding Brexit and the U . S . elections . The same approach can be used in multiple-group situations . For example , we used an in-depth attitudinal survey in Bosnia and Herzegovina ( Prism Re-
16