The Politics of Sacred Rhetoric: Absolutist Appeals and Political Persuasion

Methodological Appendix: Subjects, an Experimental Approach

215 Methodological Appendix Subjects, an Experimental Approach “If we are to talk about the human mind, let us start with human beings.” With these words, Robert Lane starts his classic text on American thought, Political Ideology: Why the Common Man Believes What He Does. Like Lane, our purpose is to understand a facet of political thinking—in this case how citizens respond to sacred rhetoric. Do human minds react to sacredness in the same way they respond to other political appeals, or does sacredness have a distinct form of influence? It should be clear immediately that this is a fundamentally empirical question. To pursue it, we cannot assume away the nature of human psychology, either in motivation or mechanics. Most of the empirical evidence from Hume to Tversky and Kahneman suggests that humans cannot be described as purely rational creatures. Moreover, the workings of the mind are more a question than a given; human psychology is illuminated by empirical study rather than logical analysis. But an empirical approach nonetheless leaves a broad range of specific methods that could be employed. Our explicit interest in this study is causation—what are the effects of hearing sacred rather than negotiable appeals? This leads us toward experimental methods, which are most suited to this task. However, this is only the case when we can manipulate the causal variable. If, for example, we thought that income explained the variation in our object of study, we could not assign different citizens to different levels of wealth in order to observe its effects. Hence we could not be sure that it is wealth rather than another factor associated with it that is really the causal agent. In these cases we turn from experiments to more inferential methods, such as the statistical analysis of survey data, attempting to control for confounding variables. In 216 g The Politics of Sacred Rhetoric our case, however, we can simply manipulate citizens’ exposure to different forms of rhetoric, establishing a clear causal test. The primary advantage of experiments is this ability to isolate causal mechanisms. This is accomplished through the establishment of a control group to compare to the treatment group, and the random assignment of subjects to one group or the other, removing selection bias and randomizing out other possible causal factors.4 These procedures establish the internal validity that is the hallmark of experimental approaches. We can be sure that any observed differences between the control and treatment groups are due to the experimental manipulation, because the two groups are otherwise identical; the only difference between them is the one that we introduced. The associated disadvantage is the lack of external validity or generalizability to circumstances outside the lab. Although experimenters strive to recreate or mimic real-world conditions, only so much can be achieved. In our case the corollary to real-world political persuasion is strong; the rhetorical statements employed are similar to actual political appeals that citizens hear in the course of their daily lives. Which brings us back to the citizens under study: our experimental subjects. The first study employs 237 students at the University of Pittsburgh , drawn from the Psychology Department subject pool, and the second sample comprises 136 students at Colgate University, drawn from the student research pool. Although the citizens in both studies are students, they represent a broader spectrum of Americans. Their representativeness relies not on the essential similarity between these students and other Americans, but on the essential similarity in how they react to political stimuli—in how their minds operate. One of the modern criticisms of laboratory experiments is the “sophomores in the lab” effect.5 The argument is that the narrow subject pool of college undergraduates may bias the results of contemporary psychology because of the unusual characteristics of that demographic, especially a susceptibility to peer pressure, less solidified attitudes , and political liberalism. Whenever any of these biases are related to the specific research question at hand, we must be especially careful about our conclusions. Of those three traits, the first is irrelevant for our manipulation, the second should not bias our specific question, and the third, if anything, biases results against our hypotheses, making the test more rigorous. Whereas some social psychology experiments regarding group dynamics and other public interactions may be biased by undergraduates’ greater susceptibility to peer pressure, our experiments are done individually, without regard Methodological Appendix f 217 to group opinion. That students...