noted below, they cannot tics is so causally complexthat honest resear - PDF document

noted below, they cannot tics is so causally complexthat honest researchers canonly make ordinal probabilisticpolitical topics,of the follow-ing form:ÒOur best judgmentis that X is less likely to betrue than Y.ÓAlan Gerber argue persuasively that the most rigorous form ofintellectual inquiry is randomized Þeld experiments. Throughoutquestions that most interest people. We cannot, alas, randomlyassign persons to democratic versus authoritarian regimes. Andinto accounts of larger, more signiÞcant political phenomena, asAdditionally, human beings are affected by the results ofthe president, that number will probably decline). Then con-sider that every stage of human scientiÞc inquiry, from ourcance?) are all inextricably normatively infused. Acknowledge,arguments pertinent to important political questions more em-overall analyses more than roughly, probabilistically true, em-pirically, and more than coherent and credible, normatively.in parts of the U.S. today, personal canvassing raises voter(Gerber and Green 2000). Yet knowing this along with other,cline in American voter turnout predominantly expresses alien-ation, satisfaction, or comparatively ineffective mobilizationtechniques. We can only say that more personalsmaller topics and it is desirable to do so. Thelar rigor on larger political matters is, however,The core value of all genuinely scientiÞc in-quiry is intellectual honesty.methods should be judged. Whether ÒtruthÓ really exists or isological standard should be intellectual honesty. We shouldthink we have learned through inquiry and analysis. We shouldobjections that might be made. Though little in human exis-tence is free of ambiguity, most of the time it is not hard towe offer well-founded challenges to misguided conventionalwe can hope to achieve on different sorts of questions. It isand therefore more scientiÞcquestions any and all evidence and arguments that can crediblysupport at least ordinal claims that ÒX is less likely than Y.Ó Political science isseen as importantand worth sustaining, other scientiÞc endeavor. We should ask instead whether itsthey really enable us to make better-founded statements aboutsults are inherently linked to peopleÕs self-understandings anddress and affect questions that they see as important for theirto do this work honestly, and to present our results in waysbilities. Put more exactly, then, I conclude that the main en-should be central to our discipline. They should not be theQuad erat demonstrandum? Admittedly, not quite. But thewww.apsanet.org Gerber, A.S., and D.P. Green. 2000. ÒThe Effects of Canvassing, Telephone Calls, and Direct Mail on Voter Turnout: AField American Political Science Review AmodiÞed version of these arguments appeared as Rogers M. Smith,Chronicle of Higher Edu-cation Sec. 2:The Chronicle Review48: B10ÐB11 (April 5, 2002).