Slide 1What is Science?What is Science?What Unites the Methods We Will Learn?What is the Perfect Research Design?Slide 6What do You Need to Do to Succeed in this CourseWhat Makes a Good Argument?What Makes a Good Argument?What Makes a Good Argument?What Makes a Good Argument?The Basics of Study DesignMake sure the project is feasibleThe Structure of ResearchTrochim’s Definitions of ValidityPOLITICAL SCIENCE 204AThad Kousser, Spring 2011What is Science?What is Science?“Science is the pursuit of knowledge and understanding of the natural and social world following a systematic methodology based on evidence.” Britain’s Science Council, 2009“Winning Arguments” – Mat McCubbins“Discovering Critical Tests” – MeWhat Unites the Methods We Will Learn?Testing a causal relationship, while ruling out alternative explanations of an empirical link.We will develop a vocabulary for different types of alternative explanations that can confound any analysis (to make it easier to think of them)We will learn a language of symbolic abstraction to summarize any research design (to make its strengths and weaknesses more clear)What is the Perfect Research Design?What do You Need to Do to Succeed in this CourseTreat it like your job; be prepared and do the workLet me get to know how brilliant you are; be ready to talk, answer questions, and anticipate questionsWrite four 5-page research design papers:Tightly writtenCreative designsDemonstrate that you get course concepts by applying all of them. (due April 26, May 10, May 19, May 31)What Makes a Good Argument?1. It is deductively validAuthors of deductive arguments contend that the truth of the premises necessitate the truth of the conclusion. It “follows from them” or can be “deduced” from them.An argument is deductively valid if the author is right, if the logic works, if the proofs check out.What Makes a Good Argument?2. It is general. Arguments that explain more phenomena are more useful, and they must explain something other than the set of facts that motivated them. 3. It has economy. Occam’s razor points us toward the simplest explanation of something, and concision often leads to generality.What Makes a Good Argument?4. Plausibility of Premises. This is the fight between those like Trochim, who want premises that are realistic, and the followers of Friedman.Friedman: “Such a theory cannot be tested by comparing its ‘assumptions’ directly with ‘reality.’ … Complete ‘realism’ is clearly unattainable, and the question whether a theory is realistic ‘enough’ can only be settled by seeing whether it yields predictions that are good enough for the purpose in hand or that are better than predictions from alternative theories.”What Makes a Good Argument?Expert billard player exampleLike the assumption of subjective expected utility maximizationUnitary actors in I.R.Mayhew’s assumption that only reelection incentive explains Congressional behaviorThe Basics of Study Design1) There must be a fully formed, clearly stated research question. 2) What do we wish to learn? Why are we conducting this research? 3) Focus is critical. There must be a clear goal in mind. Otherwise, time, energy, and money will be invested only to find that nothing has been accomplished. 2) A useful approach is to ask early on if, at the end of the project, only one question could be answered, what would that question be?Make sure the project is feasibleCan we define and measure everything that needs to be defined and measured? If the study involves some condition, can we define it? Can we be sure we'll recognize it when we see it? Democracy? Freedom House and Polity Trust in government in surveys?Using the same poll questions in different countries?The Structure of ResearchThe Structure of ResearchBegin with broad questionsnarrow down, focus in.Operationalize.OBSERVEAnalyze data.Reach conclusions.Generalize back to questions.The "hourglass" notion of researchTrochim’s Definitions of ValidityConclusion ValidityDid we find the empirical correlation that we hypothesized?Internal ValidityHave we shown cause and effect within our sample?Construct ValidityDo our measures match our concepts, or did our experimental treatment match our concept?External ValidityCan we generalize to other persons, times, or