Differences in how to do political science derive from variations among the research schools in the field. The alternative studies of turnout, which you examined in Chapter 2, and the analyses of revolution, political groups and collective action, electoral choice, and demonstrations and political violence, to which we will now turn, vary significantly in the extent to which they reflect one or more of the research schools. Recall that Piven and Cloward closely tie their work to Marx and that Riker's theory of political coalitions flows directly from the principles of rational action theory. Jackman combines rational choice theory with structural variables, and Who Votes? is even more eclectic, relying primarily on the approach of political attitudes and behavior but also drawing hypotheses from rational choice theory. Powell, as you know, insists on testing hypotheses that include variables on both the individual and structural levels of analysis. Note, too, that Ross purposefully tests hypotheses drawn from competing theories of political participation and combines the reports of anthropologists and rather sophisticated data analysis. Clearly, then, different approaches to the study of a topic produce alternative analyses.