Download presentation
Presentation is loading. Please wait.
1
PO 326 Introduction to Political Science
Week 4: Experiments & Research Ethics Core Concepts Types of Experiment: Lab; Field; Natural Research Ethics: Voluntary participation; Informed consent; Privacy Internal Validity: (Causality); External Validity (Generalizability)
2
Lecture Recap Lab Experiments: Experiments conducted in a facility that provides controlled conditions (See Essex University Lab) Field Experiments: Experiments which take place in real-world settings Natural Experiments: Experiments in conditions which occur naturally, in so far as the researcher is not active in the data gathering process
3
Basic Principles of Experimental Design:
1. Two groups: One group that is exposed to the intervention (the experimental group) and one group that is not exposed to the intervention (the control group) 2. Random allocation of subjects to the groups before the pre-test 3. One pre-intervention (pre-test) measure on the outcome variable 4. One intervention (test/treatment) 5. One post-intervention (post-test) measure on the outcome variable (see de Vaus 2001: 48) (Also see H & H: pp ) (See de Vaus 2001: 48)
4
Why conduct Experiments?
Researchers can investigate causal hypotheses with a great deal of confidence- i.e. X---> Y. Traditional survey and correlational research designs cannot show causation empirically. Practical and has policy relevance (Stoker, 2010) Academics can collect data to test hypotheses that they would otherwise not have been able to investigate Valid and reliable evidence
5
Examples: Social Experiments and Poverty
Solution: Random Treatment Assignment Political Mobilization Coalition Bargaining Effect of Voting Ballots & Turnout Foreign Policy Decision Making And many more!!!
6
Experiments If causal inference (Internal Validity) is the crucial methodological problem for social scientific research, then experiments are the ideal solution. Discuss.
7
Experiments A researcher randomly selects the names of 5000 people from the electoral register of a constituency in Edinburgh and randomly allocated them into two groups (control and experimental). S/he sent letters to the experimental group encouraging them to vote in the elections, but did not contact the control group. What type of experiment did s/he conduct: natural, laboratory or field? Why do we need ‘control groups’ while conducting experiments?
8
Experiments Which questions might best be best suited for experimental research?
9
Milgram‘s Experiment S. Milgram (1963): Behavioral Study of Obedience
Participants: 40 males (age 20-50) Participants were told they are participating in a study regarding the relationship between punishment and learning Set up: “teacher” and “learner” Punishment is administered by the “teacher” through electric shocks generator had 30 switches labelled with voltage levels, ranging from 15 to 450
10
Milgram‘s Experiment Operational definitions
Defiance: any subject who stopped the experiment at any point before the 30th shock level Obedience: any subject who complied with all commands and administers all shocks.
11
Milgram‘s Experiment Results
26 of the 40 participants obeyed till the end, administering 450-volts to the learner 14 subjects defied the experimenter No subject stopped before 300 volts 5 of the 40 participants stopped at 300-volts 4 stopped at 315-volts 2 stopped at 330-volts 1 stopped at 345, 1 stopped at 360 1 stopped at 375-volts
12
The Milgram Experiment
13
Research Ethics How does this study challenge our notion of “research ethics”?
14
Research Ethics One criticism of Milgram’s experiment is that the participants were not appropriately debriefed and supported afterwards (many of them struggled with knowing themselves capable of great harm). Is debriefing enough? Is deception unethical in social science experiments?
15
Research Ethics Milgram (1964) responding to the criticisms of his study surveyed his participants after debriefing them and found that 83.7% were glad to have participated and only 1.3% regretted it.
16
Additional Group Work You want to test whether commercials have an effect on people’s propensity to start or quit smoking What type of experiment would you conduct? How would you design this experiment? What are the advantages and limitations of your experiment?
17
Seminar Recap Applications of Experiments in the Social Sciences
Advantages and disadvantages to different types of Experiments Example: The Milgram Obedience to Authority Experiment. For further reading, see the Stanford Prison Experiment (1971) Deadlines: Research Report (3/11/14) by Week 6
Similar presentations
© 2025 SlidePlayer.com. Inc.
All rights reserved.