Chapter 7: Audio Experiments - What Causes What?

Key Topics

• Logic of experimentation
• Variations in experimental design and context
• Process of planning and conducting an experiment
• Strengths and weaknesses of experiments

Example Study: Prosecutorial Misconduct

• Hypothesis: Greater severity of a crime increases likelihood of prosecutorial misconduct.
• Experiment Design: Laboratory experiment with undergrad participants acting as prosecutors.
  • Random assignment to murder or assault case.
  • Manipulation of the independent variable (severity of crime).
  • Measurement of prosecutorial misconduct as the dependent variable.

Logic of Experimentation

• Causal Inference: Establishing a causal relationship between variables.
• Association: Differences in dependent variables between experimental conditions.
• Direction of Influence: Time order of events; dependent variable measured after manipulation.
• Non-spuriousness: Conditions differ only by manipulation; random assignment maintains equal groups.

Random Sampling vs. Random Assignment

• Random Sampling: Selecting a representative sample from the population.
• Random Assignment: Assigning sampled participants to different groups within the experiment to ensure equivalence.

Variations in Experimental Design

• Timing of Measurements: Pre-test, post-test, or both.
• Factorial Designs: Study of two or more independent variables.

Variations in Experimental Context

• Laboratory vs. Field Experiments: Laboratory offers control; field experiments offer real-world applicability and increased external validity.
• Embedded in Surveys: Can test things like question wording within surveys.

Process of Planning and Conducting Experiments

1. Design the Experiment: Develop hypotheses and operationalize variables.
2. Pretest: Conduct a trial run to identify potential issues.
3. Recruit Participants: Use incentives if necessary.
4. Obtain Informed Consent: Explain the study and participant rights.
5. Randomly Assign Participants: Ensure balance between groups.
6. Manipulate the Independent Variable: Set different criteria.
7. Measure the Dependent Variable: Check for expected changes.
8. Debriefing: Explain any deception and gather participant feedback.

Strengths of Experiments

• Evidence of Causal Relationships: High internal validity.
• Controlled Environment: Allows manipulation of variables without external interference.

Weaknesses of Experiments

• Low External Validity: Results may not generalize outside the lab.
  • Counteracting Measures: Field studies, replication, increasing sample size.
• Reactive Measurement Effects: Participants may alter behavior when observed.
  • Counteracting Measures: Blind or double-blind studies, careful debriefing.
• Topic Limitations: Not all subjects are suitable for experimental research.

Conclusion

Experiments are powerful for establishing causal relationships but have limitations due to control environments and potential lack of generalizability. Balancing strengths and addressing weaknesses is key for effective experimental research design.