POS3713 Midterm 2 Study Guide and Notes POS3713 Midterm 2 Study 1 What is a research design and what are its chief goals a Research Design i Formally 1 Specifies the procedures we will use to answer a research question a What date will we use Time Space b What hypothesis tests most appropriate 1 Seeks to ensure that data we examine provide an answer to research question ii Informally b Goals i Goal Allow for valid inferences as to causal effects of x on y ii Good research design will account for 1 Threats to causality 2 Unit of interest 3 Temporal spatial variation of interest c All research designs specify comparisons i Comparison fundamental building block for causal inference 1 May compare values of Y across units 2 May compare values of Y over time ii Goal Make comparisons as similar as possible with exception d Goal of Research Design Compare like with like i To assess whether X causes Y we must compare two or more of X units ii Ex We suspect that rainfall X on election day influences voter turnout Y by precinct iii We must compare rainfall and turnout in one precinct to another precinct where rainfall is different 2 What is an experiment How does it differ from an observational study a An experimental design i Researcher randomly assigns subjects to one of two groups the experimental group take drug and the control group placebo ii Researcher control the introduction of the experimental factor the drug blood pressure iii Measure the two groups on the dependent variable of interest iv Compare the two groups If there is a difference you can attribute it to the experimental factor the drug b An experiment controls and randomly assigns X i A researcher controls X using random assignment ii Observational is any other design iii Control 1 Value of the IV determined only by researcher iv Random assignment makes it different from observational study IV 1 Researcher must randomize assignment of the value of a Participants randomly assigned to treatment or control groups b Random assignment helps ensure all groups are identical on all Z variables 3 What are the strengths and weaknesses of experiments Observational studies a Experiments i Not every independent variable X is controllable and subject to experimental manipulation 1 Gender 2 Religion 3 Political Party ii Carry special ethical dilemmas for the researcher 1 Milgram experiment where the subject shocked people 2 You couldn t test certain cancer treatments and their survival rates iii We sometimes make mistakes of emphasis when interpreting the results of the experiment 1 If an experiment produces a finding that some X does indeed cause Y that does not mean that that particular X is the most prominent cause of Y iv Experiments have high internal validity often have low external validity 1 Internal validity a Confidence that the effect of X on Y is real b We have crossed the four hurdles of causality 2 External validity a Experiments do not require a random sample of b the target population It is extremely rare for experiments to draw a random sample c College students are recruited from colleges a sample of convenience d Scientists will generally use replication to solve this issue to see if their results apply to a broader population 3 Threats to internal validity a Any differences between participants in treatment and control groups other than X 4 Threats to external validity a Any differences between i Participants and non participants ii Setting under study and real world b Observational Studies i Observational Studies have low internal validity often have high external validity ii 1 Internal validity The Colbert Bump 1 Level of charisma of candidate 2 Experience gender 3 Candidate s policy views iii 2 External validity The Colbert Bump 1 Sample size unrepresentative 2 Differences between those under study and the population we care about iv Overall which do you think is the bigger threat 1 In most cases experiments internal is good and external is weak 2 Observational internal is weak and external is better v When we need to control for the other possible causes of Y to cross the fourth causal hurdle we need to control for all of them not just one and with observational studies we do not know for certain if we have vi Observational analysis must be a bit more tentative in its pronouncements about causality 4 What are field experiments and how do they compare to traditional experiments and observational studies a Field Experiments one that occurs in a natural setting where the subjects normally lead their lives Random assignment to treatment groups has enabled researchers in the social sciences to study subjects that seemed beyond their reach for experimentation i Advantages ii Disadvantages 1 Natural setting high external validity 2 Longer observation periods 1 Do not have complete experimental control 2 May not be able to use random assignment a Therefore treatment and control groups not balanced on Zs iii Different from observational study because experimenter is trying to assign x with a randomization process and influence it iv Different from experiment because it takes place in a natural setting harder to control v Example Get out the vote 1 Do phone calls increase turnout X non partisan telephone appeal to vote a b Y vote 2 Results a 48 turnout for control b 65 for treatment 3 Sociability difference might not answer the phone and might not want to interact with people and vote 4 Not an experiment because cannot control for all variables a Lack of experimental control 5 Placebo telephone appeal to donate blood 6 Results a Placebo 67 turnout b Treatment 65 turnout 7 Conclusion a Phone calls don t increase turnout 5 What is the difference between random assignment and random selection What types of studies use each How does each contribute to internal or external validity a Do not confuse random assignment with random selection i Random assignment values of IV assigned randomly experiments high internal validity ii Random selection members of study selected randomly from population observational studies high external validity 6 What are the two types of observational studies Why would a researcher choose to use each one a Good for situations where you cannot control for X gender religion political party b Cross Sectional Observational Studies i Examines a cross section of social reality focusing on variation between individual spatial units like citizens elected officials countries and explaining the variation in the
View Full Document
Unlocking...