POS3713 Without the Scientific Method we could only describe political phenomena the Who and What The Scientific Method allows for casual inference the Why and How Epistemology the study of what we know and how we know it Traditionally we know through 1 Recognized authorities 2 Precedent 3 Intuition 4 Common Sense 5 Observation Observing the Empirical World Verifiable and Repeatable You can constantly observe phenomena occurring Falsification Where we observe something x influencing something else y but you can also observe x not effecting y Probability We sometimes see x effecting y and sometimes we do not X influences y with a certain probability When x is present y occurs 75 of the time Moving From Observation to General Theory Political Science in rooted in observation We take an observation and then generalize to a more broad population As an example Your family education and turnout Do these apply beyond your social group The goal of scientific research is inference Inference A conclusion reached on the basis of evidence and reasoning Features of Scientific Research We create an interesting plausible story to answer a question of interest We use transparent methods to investigate Assess how confident you are in your explanation ie how much uncertainty is there Science is the process for determining if your explanation is worth while Inference Uncertainty We want to infer a relationship between inference and uncertainty which requires qualitative and quantitative judgements How we make inferences 1 We cannot observe causation we do not see X make Y happen 2 We observe correlation thus seeing X and Y move together 3 We then come up with a Causal Story to explain why the correlation take place Prediction using information to predict the future value of a concept or variable Correlation where two variables seem to move together Causation where change in one variable leads to change in another variable this requires correlation Correlation Causation Deterministic Relationships When x occurs y will occur with certainty Cause x usually leads to effect y Probabilistic Relationships When x occurs y will occur with certainty Cause x usually leads to effect y There is an element of uncertainty with both models Science seeks to advance knowledge and provides a process by which we generate knowledge Where does knowledge come from 1 Mutual Agreement 2 Daily Experience 3 Tradition 4 Authority Logic and Empirics Science provides a systematic method of knowing We cannot simply observe things and know Why not 1 2 Selective Observation 3 Illogical Reasoning 4 Over generalizing Inaccurate Observations Public Criteria for Science Evidence and Reason Empirical facts should support your explanation Emotion or passion cannot overrule evidence or reason Fallacy of Affirming the Consequences All you can say is that you have yet to falsify your theory If A is true then B is true this does not hold logically I observed B therefore A may be true If x causes y then x and y are correlated From Science to Political Science The world is comprised of data Data is phenomena occurring Theories Hypothesis Proposed relationships must be testable and you must know what evidence is relevant for consideration Hypothesis the relationship we expect to observe in our data The hypothesis moves us from the Theoretical level to the Empirical level The hypothesis is a testable statement derived from the theory that indicates cause effect within the data A positive relationship within data is signified by x and y moving together A negative relationship within the data is signified by x and y moving apart We want the testable hypothesis to Make strong and explicit propositions Specify the existence of a relationship between the dependant and independent variables Specify the nature of that relationship 1 Positive or Negative 2 Linear or Curvilinear Requirements for a Hypothesis 1 Must offer a cause and effect 2 Must specify the direction whether it is positive or negative 3 Must specify the type of relationship whether linear or curvilinear 4 Must specify units variation of interest such as Units individuals or aggregates Variation Over time or between units Two Negatives is a Positive The Null Hypothesis the expected relationship if our theory is wrong implicit in all hypothesis When we test our hypothesis we are trying to reject our null hypothesis If we cannot say that our theory is wrong then we can believe that our theory is right Falsifiablilty The ability of a hypothesis to fail to be falsified helps to discern how worthwhile a theory is If you can falsify the null hypothesis your theory is relatively safe but you should continue testing There is no number of set tests that a theory must pass before we can accept it Even if a hypothesis is falsified does not mean that we have to abandon a theory Perhaps if there were better measures involved in the study out theory would be supported Steps in Testing our Hypothesis We must make our concepts operational First we need a solid theoretical definition abstract definitions are neither true nor false they are explicit and used for the purpose of clearly stating your theory Operational definitions explain what concepts will look like and how it is to be taped and measured in the empirical world always specify unit of measurement Evaluating Operational Definitions Validity How well does the operational definition match the theoretical definition Reliability How consistent will the process by which we measure these concepts be There is always the risk of measurement error whether random or systematic Identify a Question or Puzzle Identify the Independent and Dependant Variables Summing Up the Method 1 2 3 Propose a Causal Theory the story that links the variables 4 State the Hypothesis also identify the Null Hypothesis 5 Operationalize the Variables meaning measure them and turn them into numbers 6 Test the Hypothesis Before the Theory Observe or Identify a Puzzle or Theory The goal is to generalize What is the general phenomena you are looking to explain The primary interest is to explain change or variation in the dependant variable Identify the time dimension At what point in time are we measuring the variables Then identify the spatial dimension what units we want to measure Time Series Analysis same units measured at different points in time Spatial Variation Different units measured at one point in time Building A Theory You need Actors or Units of Theory What
View Full Document
Unlocking...