|
Abstract:
|
Political change in settings of weak institutions can occur in a wide variety of ways. Individuals might believe that the best way to create political change is through traditional means, such as voting, or through unconventional means such as protests or even armed action. Dorff developed a data set with demographic, biographic, and attitudinal responses from approximately 1000 individuals from thirty federated states and the Federal District of Mexico. The responses include a categorical choice of strategy for political change. Multinomial logistic regression of the change strategy response on the respondent's characteristics was performed by aggregating the states into regions as fixed effects and also by using Bayesian analysis to fit a mixed model with the individual state of residence of the respondent treated as a random effect. These analyses showed statistically meaningful associations of interest to professionals in International Studies. Graphical methods based on the strengths of the two approaches are used to make the associations and their uncertainties accessible to the wider International Studies audience.
|
