STUDY GUIDE ON POS3713 FINAL (EVERYTHING YOU NEED TO KNOW)UNIT 1Epistemology—the study of what do we know and how do we know it. How we believe something is true. Knowing does not equal truth.Inference—the goal of scientific research is inference. Descriptive inference is using observations about the world to learn about other unobserved facts. Causal inference is cause and effect assumption. We want to infer a relationship between two concepts—this requires quantitative and qualitative judgments. X causing Y. WE cannot observe causation, we don’t see x make y happen. We observe correlation in which we see x and y move together. We come up with a causal story to explain why the correlation takes places—x and y occurs together because x causes y—and the story is the why and the how x causes y. Deterministic Relationship—If some cause occurs, then the effect will occur with certainty (physical sciences world). Cause X always lead to effect Y; when x occurs, y will also occur with certainty. E.g. Force= mass x acceleration. The effect is all the time guaranteed. Probabilistic or Stochastic Relationship—cause X usually leads to affect Y. When X occurs, Y will tend to occur, but not with certainty. Increases in X are associated with increases (or decreases) in the probability of Y occurring, but those probabilities are not certainties (world of human interactions). Stochastic in meaning the randomness in the world, the social world is stochastic. IN our probabilistic models the best we can do is say that x appears to increase (or decrease) the probability of y occurring or changing. These models cannot prove causality. The key to inferring causality is research design. Previous Research and Theories—without scientific method, we can only describe political phenomena, the who and the what. The scientific method allows for causal inference, the why and the how. Political science is rooted in observation. Duberger’s Law—A principle which asserts that a majority voting election system naturally leads to a two party systemHotelling’s Law—In many markets it is rational for producers to make their products as similar as possible. When examing previous research: must have a skeptical view of research and theory. Must think about these questions—what if any other causes of the dependent variable did the previous miss? Can their theory be applied elsewhere? If we believe their findings, are there further implications? How might thistheory work at different levels of aggregation (micro<->macro)?Conceptual Definitions—the meanings we assign to terms. Concepts are a shared understanding. To have a clear sense of what the concept is that we are trying to measure and we are all using the same definition for all language.Independent and Dependent Variable—independent variable is the single cause. Dependent variable is the single effect. The value of the “dependent” variable “depends” on the value of the “independent” variable. According to our theory, a change in the value of the independent variable causes change in the value of the dependent variable. Correlation and Causation—Correlation—2 variables seem to “move together” or “apart”; Causation—changes in one variable lead to changes in another variable (this would require correlation)—correlationdoes not equal causation. Correlation is necessary but not sufficient for inferring causation. Correlation can be called covary. It is possible for a causal relationship to exist between X and Y even if there is not bivariate association between X and Y. 4 hurdles when establishing causal relationships: 1. Credible causal mechanism connect X and Y. The how and why of relationship. 2.Ask whether it is actually possible (or even likely) that Y might cause X. 3.Consider whether X and Y are correlated—simple bivariate relationship. 4. Consider a confounding variable “Z” that is related to both X and Y and makes the association between X and Y spurious. The Fallacy of Affirming the Consequent—We have evidence and reason in our favor, therefore, we have proved our theory but in reality all we can say is we have yet to falsify our theory. Formal fallacy, always a bad argument, it is affirming the consequent. Example:1. If I am lecturing, then I am standing. If A, then B2. I am standing----------------------------- B3. Therefore, I am lecturing Therefore ATime Series vs. Cross Sectional—Time series: Temporal variation (time dimension) identifies the point orpoints in time at which we would like to measure our variable. The spatial dimension in a time series measure is the same for all cases and the dependent variable is measured at multiple points in time. Cross Sectional: Spatial dimension identifies the units that we want to measure. Cross sectional measures is when the time dimension is the same for all cases and the dependent variable is measured for multiple spatial units. Measuring our dependent variable such that one of these 2 dimensions will be static (or consistent). This means that our measures of our dependent variable will be of one of two types. A time series example includes: the average monthly level of U.S. presidential approval displayed from 1995-2005. Spatial unit is same (US) but variable has been measured at multiple points in time (each month).Example of time seriesCross sectional example: military spending as a percentage of gross domestic product (GDP) in 2005 for 24 randomly selected nations.Starting with a puzzle—2 cases expected to have similar outcomes, but they don’t OR 2 cases expected to have different outcomes but they do. Means of Evaluating Theories (Strengthening Theories, Generalization, falsifiable, parsimony)—1. Falsification—we observe something (x) influencing something else (y), we may also observe when x does not influence y—falsifiable theory—there has to be some way the theory can be disproven (i.e. evidence). 2. Generalization—theory should pursue generality, meaning we want our theories to be applied to as general a class of phenomena as possible. We take an observation and then generalize to a broader population. Example: theory explaining phenomena across multiple countries is stronger than theory explaining in only one country. Being able to use similarities to apply theories to other conditions or subjects3. Parsimony—meaning theory should be simpler, which causes it to be
View Full Document
Unlocking...