Notes on Factorial Designs
Factorial Designs
- Definition: Factorial designs involve experiments with two or more independent variables (IVs). They allow researchers to observe interactions and main effects of the independent variables.
Research Goals and Claims
- Describe behaviour: Claim frequency of behaviours (e.g., prevalence of depression).
- Predict behaviour: Establish associations (e.g., relation between depression and social isolation).
- Explain behaviour: Make causal claims (e.g., social isolation leading to depression).
Simple Experiments
- Independent Variable (IV): Manipulated to examine effects on a dependent variable (DV).
- Example 1: Treatment vs. Control on Materialism.
- Example 2: Exercise vs. No exercise on depression levels.
Adding Levels to an IV
- Increasing IV levels provides detailed insights:
- Can identify boundary conditions and curvilinearity.
- Example: Hours of exercise and its relationship to depression (0, 1-2, 3-5, 6+ hours).
Reasons to Add Levels to an IV
- Identify Boundary Conditions: Understand how different conditions affect the outcome.
- Curvilinearity: Analyze nonlinear relationships.
- Example: Anxiety levels affecting test performance.
- Test Multiple Treatments: Compare various methods or interventions.
Costs of Adding Levels to an IV
- Power and Error Rates: Increase in Type II errors due to the larger sample size requirements.
- Resource Needs: More complex designs require more resources.
Factorial Designs Overview
- Factorial designs cross two or more IVs:
- Example: Effects of cell phone use and age on driving performance.
- Participant Variables: Selected/measured characteristics (e.g., age, gender).
Testing Theories with Factorial Designs
- Can test boundary conditions and generalizability of effects:
- Does IV impact different groups the same way? (e.g., alcohol’s effect on aggression varies with body weight).
- Testing whether effects generalize across different cultures.
Interpreting Effects in Factorial Designs
- Main Effect: Overall effect of one IV on the DV, averaging over other IVs.
- Interaction Effect: Effect of one IV depends on the level of another IV.
Examples:
- Study of well-being with cats vs. dogs illustrates marginal means and main effects.
Complex Factorial Designs
- Can include more IVs beyond simple 2x2 layouts (e.g., 2x2x2, 2x4x2).
- Complexity increases power and sample size requirements.
Practice Questions
- Apply knowledge of designs and effects to hypothetical studies.
- Example: Determine main effects and interactions in a study comparing re-reading vs. flashcards across subjects (History and English).