Exam 1 Review Sheet Public Opinion Overview Definition of Public Opinion Consensus definition for class 1 Brooker and Schaefer The Expressed attitudes and views of ordinary people on issues of 2 V O Key Jr Those opinions held by private persons which governments find it prudent to 3 Clawson and Oxley The sum or aggregation of private opinions on any particular issue or set public concern heed of issues Classical vs Representative Democracy Definition and justifications for each 1 Classical Democracy The Greek Model a High Levels of interest in public matters b Magnanimity rather than selfishness c The public has a say in all policy decisions 2 Representative Democracy James Madison via John Locke a Citizens are not capable of what is required of them i Not knowledgeable enough ii Too self interested b Direct Democracy not feasible in a huge country c Similar to elitist and pluralist conceptions Normative Democratic Theories Participatory Elitist and Pluralist definitions and examples 1 Participatory Emphasizes citizen involvement more than other alternatives 2 Elitist Democratic Elitism a Elections the mechanism for the public to express preferences Is this more viable b Politicians are accountable to the public but c Indirectly so 3 Pluralist The preferences of the public are expressed through groups i Citizens join groups that represent their specific interests Famous polling failures why 1 Problems with the Literary Digest Poll a Sampled motor vehicle registration lists b Telephone books c Response bias Key terms in public opinion beliefs values emotions 1 Belief a feeling that some is true 2 Values A particular kind of belief global principles of how important a generalized concept is a Types of beliefs i Factual Information ii Stereotypes to a person 3 Emotions Feelings that a political object creates a Can be positive happiness elation or negative anger fear b Not always subject to rational change Meaning of social desirability bias 1 Social Desirability Bias a Survey respondents want to please the interviewer Potential polling biases discussed in class selection self selection response order wording order environmental i Selection When a random sample isn t really random ii Self Selection The variable of interest might not be randomly assigned 1 Test preparation classes and test scores Political knowledge and education iii Response Order Sometimes people choose the first answer they encounter that seems reasonable or simply click the first answer to all questions because they are lazy tired iv Question wording order of question environmental factors Implications of Bartels 2003 i Americans don t care about income inequality Statistics Experiments in Political Science Sampling types and definitions of each 1 Quota Sampling Try to create a sample that looks like the public at large Used by Gallup in his original surveys Interviewed a certain number of people from every location income gender age and race A good alternative if random sampling is not viable 2 Convenience Sample Examples O Reilly Poll Literary Digest FSU Study Problems o Lack external validity o Unlikely to be representative of the broader population o Those that select in may be special eg O Reilly 3 Simple Random Sample o The most common type of sampling in public opinion o A sample of respondents is drawn at random from the population of interest o The Gold Standard in public opinion research 4 Longitudinal Conducted over long period of time Confidence interval and margin of error Means and medians when to use each Meaning and implications of survey weighting 1 Confidence Interval The probability that a random sample drawn from the population will fall within a certain range In Statistics we most often want to find a 95 confidence interval 2 3 Margin of error The values of a particular variable that cover in theory 95 of random samples of that variable in the population 4 Means are appropriate when data follows a normal distribution 5 When data is skewed report the median 6 Survey Weighting Increasing or decreasing the importance of a respondent in a survey based on a comparison of the demographics of that respondent to the population of interest a Over represented groups receive weights less than 1 b Under represented groups receive weights greater than 1 Push Polls Establishing causality 3 components 1 Causality A theory that X results from Y a Difficult to establish in non laboratory studies i Myriad events occur in the real world 2 To establish that X causes Y you must show that a X comes before Y Temporal Precedence b Y varies with X Covariation If X then Y i ii If not X not Y c No Plausible alternative explanations Endogeneity definition 1 Endogeneity a How do we know that x caused y and y didn t cause x b A failure of temporal precedence Types of experiments classic post test multi group 1 Classic The key dependent variable is measured before and after treatment a Comparisons can be made on change within subjects b Generally ideal 2 Post test The key dependent variable is only measured after treatment a Comparisons are made on change between subjects 3 Multi group There are multiple treatment groups and at least one control group Experimental venues validity lab field natural 1 Laboratory Experiments a Like the FSU studies b Experimenters have almost total control of the environment and access to stimuli i Highest possible internal validity 2 Field Experiment a Conducted outside of the lab on some subset of the population b Experimenters still control exposure to a stimulus but have a lower level of control over the environment 3 Natural Experiments experiment c Mid level internal validity higher external validity Researchers take advantage of an event in the real world that functions like an Experimenters have little no control over exposure to the stimulus o Rely on natural randomization often problematic o Highest external validity lowest internal validity Experiments in various social sciences economic psychological political science 1 Economic Experiments Use monetary or other incentives to induce rational behavior a Test the viability of the rational man b Like the Grosser and Schram 2010 article 2 Psychological Experiments Utilize deception less likely to utilize incentives Deception is thought reasonable so long as it is necessary for the success of the study 3 Political Science A hybrid of the economic and psychological approaches o The drive for causality in political science o Not popular