Methodology: How Social Psychologists Do Research
Social Psychology as an Empirical Science
- Goals of Social Psychology:
- Describe behavior
- Explain behavior
- Predict behavior
- Methods of Inquiry:
- Correlational Method: Assesses relationships between variables
- Experimental Method: Determines causality through controlled experiments
- Meta-Analysis: Summarizes findings from multiple studies
Understanding Predictions in Social Experiments
- Hindsight Bias: Tendency to see outcomes as obvious after they have occurred.
- Example: Predicting election results becomes easier post-election.
- Hypotheses often arise from everyday observations.
- Case Study: Kitty Genovese's murder led to the hypothesis of Diffusion of Responsibility by Latané and Darley, where individuals assume others will take action.
- Research is essential to test these hypotheses: mere speculation is not sufficient.
Primary Research Designs
- Correlational: Measures how two or more variables are related.
- Experimental: Includes random assignment and control to identify causal effects.
- Meta-analysis: Compiles data from previous studies for broader insights.
The Correlational Method
- Positive Correlation: Both variables increase together (e.g., height and weight).
- Negative Correlation: One variable increases while the other decreases (e.g., job satisfaction and quitting thoughts).
- Correlation Coefficients: Ranged from -1.00 (perfect negative) to +1.00 (perfect positive), with 0 indicating no correlation.
Limitations of Correlational Method
- Correlation does not imply causation. Three possible causal relationships may exist:
- Variable A causes B
- Variable B causes A
- A third variable influences both A and B.
The Experimental Method
- Experimental Design:
- Participants randomly assigned to conditions to determine causality based on independent variable manipulation.
- Independent Variable: The variable manipulated by researchers (e.g., type of TV program watched).
- Dependent Variable: The outcome measured to assess the impact of the independent variable (e.g., aggression levels).
Validity in Experiments
- Internal Validity: Ensures the independent variable is the only thing affecting the dependent variable, maintained through control of extraneous variables and random assignment.
- External Validity: Refers to the generalizability of findings beyond the study. It addresses:
- Generalizability across situations
- Generalizability across people
Enhancing External Validity
- Field Experiments: Conducted in real-world settings to enhance ecological validity while lowering control compared to lab experiments.
- Realism Types:
- Mundane Realism: Similarity of the experiment to everyday situations.
- Psychological Realism: Similarity of psychological processes in the experiment to real-life situations.
The Replication Crisis
- Issues with replicating landmark studies due to:
- Small effect sizes
- Selective publication and biases in research practices.
- Meta-Analysis aids understanding of the reliability of effects across studies.
Ethical Considerations in Psychology
- Researcher Integrity: Need for ethical standards to avoid data fabrication, suppression of results, or undue pressure on participants.
- Participant Welfare: Protecting participants from stress and ensuring informed consent are key ethical responsibilities.