Research Designs in Psychology

Goals of Different Research Designs

• Descriptive Research: Provides a snapshot of the current state of affairs.
  • Advantages: Offers a complete picture of what is occurring at a given time; helps in developing questions for further study.
  • Disadvantages: Does not assess relationships among variables and cannot infer cause and effect.
• Correlational Research: Assesses relationships between two or more variables.
  • Advantages: Tests expected relationships and makes predictions; applicable to everyday life events.
  • Disadvantages: Cannot infer cause and effect.
• Experimental Research: Evaluates the causal impact of experimental manipulations on a dependent variable.
  • Advantages: Enables conclusions about causal relationships.
  • Disadvantages: Many important variables cannot be manipulated; may be costly and time-consuming.

Descriptive Research: Assessing Current State

• Case Studies: Detailed records of individual experiences and behavior.
  • Example: Freud's analysis of Little Hans; Phineas Gage’s study.
  • Challenges: Limited generalizability; potential inaccuracy or incompleteness.
• Surveys and Tests: Gather beliefs or behaviors of a sample.
  • Validity: Ensures the instrument measures what it's supposed to.
  • Reliability: Consistent results over time or situations.
  • WEIRD Samples: Overrepresentation of Western, educated, industrialized, rich, and democratic populations.
• Naturalistic Observation: Observing behavior in its natural environment.
  • Laboratory Observation: Offers more control but may lack authenticity.

Correlational Research: Relationships Among Variables

• Correlation Coefficients: Measure the association; ranges from -1 to +1.
  • Positive correlations: Higher scores on both variables.
  • Negative correlations: High score on one variable, low on the other.
• Scatterplots: Visual representation of the relationship.
• Three Pathways of Cause and Effect:
  1. X causes Y.
  2. Y causes X.
  3. Third variable (C) causes both X and Y.
• Strengths: Useful when experiments aren't possible; studies behavior in everyday settings.
• Limitations: Cannot determine causal relationships.

Experimental Research: Causal Relationships

• Independent Variable: Manipulated by researchers.
• Dependent Variable: Outcome affected by the independent variable.
• Key Criteria:
  • Random assignment to conditions.
  • Control condition for comparison.
  • Large and diverse sample size.
  • Minimized experimenter effects.
• Confound: An extraneous variable affecting results.
• Limitations: Laboratory settings may not mimic real life; some variables cannot be manipulated.

Examples

• Video Games and Aggression Experiment: Random assignment to violent vs. non-violent games; measured aggressive behavior.
  • Demonstrates initial equivalence and causal inference.
• Potential Confounds: Variables that interfere with the independent variable's effect.