Chapter 12

Research Methods: A Process of Inquiry - Chapter 12 Factorial Designs

General Overview

  • Focus on Factorial Designs: A methodology that involves studying two or more independent variables simultaneously.

  • Significance: Allows researchers to investigate not just the main effects of each independent variable but also the interactions between them, which is critical in psychology where variables frequently interact.

Key Components of Factorial Designs

  • Definition of Factorial Designs: Designs that include two or more independent variables.
      - Essentially understood as conducting two or more experimental studies within a single framework.

  • Purpose of Factorial Designs: To explore the interactive effects of multiple independent variables on a dependent variable.
      - Example: A study examining the impact of study environment (quiet vs. noisy) and study method (highlighting vs. summarizing) on student performance.

  • Main Point: Most independent variables in psychology tend to interact with others, making factorial designs particularly valuable.

Main Effects and Interactions

  • Main Effects: The individual effect of each independent variable on the dependent variable.

  • Interactions: Occurs when the effect of one independent variable differs based on the level of another independent variable.
      - Definition: A combined effect that is not merely the sum of the main effects.
      - Example: The impact of study method may significantly differ when combined with different levels of noise.

Graphing Factorial Designs

  • Steps for Graphing: For designs with two independent variables:
      1. Choose one independent variable for the X-axis and label it accordingly with its levels.
      2. Label the Y-axis to sufficiently display the mean scores from each cell (the combinations of levels from both variables).
      3. Graph mean results for the first level of the second independent variable, labeling that line. Repeat for each level of the independent variable.

  • Sample Data to Graph:
      - Structure:
        - A1: B1 = 10, B2 = 20
        - A2: B1 = 20, B2 = 40
        - A3: B1 = 50, B2 = 40

  • The resulting data points must be accurately represented in terms of their respective means and levels.

Example Study: Children’s Dark Fears Study

  • Factors:
      - Factor A (Illumination): Levels include Lighted and Dark.
      - Factor B (Images): Levels include Frightening and Neutral.

  • Hypothesis: Fear of darkness in children may combine with frightening visuals to exacerbate their fears.

Factorial Design Logic

  • Matrix Representation: This aids in understanding how main effects and interaction effects are formulated.
      - Main Effects for Factor A: Comparing mean scores across levels of illumination.
      - Main Effects for Factor B: Similarly comparing mean scores across levels of image type.
      - Testing Interaction Effects: Conducting an analysis on how different combinations (A1B1, A1B2, etc.) influence outcomes.

ANOVA Summary Table for the Study

  • Provides details on factors, degrees of freedom (df), sum of squares (SS), mean square (MS), F-ratio, and p-values:
      | Source | df | SS | MS | F | p |
      |--------|----|----------|----------|------|------|
      | Factor A (illumination) | 1 | 765.62 | 765.62 | 7.88 | .008 |
      | Factor B (images) | 1 | 525.62 | 525.62 | 5.41 | .026 |
      | AB Interaction | 1 | 497.02 | 497.02 | 5.12 | .030 |
      | Error | 36 | 3497.50 | 97.15 | | |
      | Total | 39 | 5285.78 | | | |

Evaluating Main Effects

  • Process: Examine all participants tested under different conditions of Factor A, irrespective of Factor B levels, followed by similar evaluation for Factor B.

Evaluating Interactions

  • Graphical Analysis: Is crucial for demonstrating interaction effects. Non-parallel lines on a graph indicate significant interaction.

Possible Outcomes in Factorial Designs

  • Outcomes demonstrate various interactions and main effects. Key potential configurations:
      - 2 × 2 factorial, where none of A, B, nor their interaction are significant.
      - Variations indicating significance for A or B or their interaction, impacting subsequent interpretations of results.

  • Combinations of significance illustrate the complexity and richness of insights obtainable through factorial designs.

Factorial ANOVA

  • ANOVA is applicable to analyze various factorial designs.
      - The number of effects recognized correlates directly with the number of independent variables under consideration:
        - 2 IVs: Examine factors A & B and their interaction.
        - 3 IVs: Examine main effects A, B, C and interactions therein (including AB, AC, BC, ABC).
        - 4 or more IVs: The complexity escalates significantly, potentially requiring advanced statistical expertise.

Repeated Measures Factorials

  • Concept: Uses a within-subjects design where participants undergo each condition, warranting control of sequence effects.
      - Illustration: A1 -> A2, indicating sequential conditions evaluated in repeated measures.

Mixed Designs

  • Definition: Combine different types of independent variables, such as manipulating certain factors while others remain non-manipulated.
      - Example: A mixed design with distraction levels as within-subjects and reward amount as a between-subjects factor.

Variations on ANOVA

  • ANOVA can adapt to various independent variables and factor combinations, leading to diverse analyses:
      - ANCOVA: Analyze with covariates included.
      - MANOVA: Handle scenarios involving multiple dependent variables.

Group Exercise Discussion Points

  • When are factorial designs most beneficial?

  • Explain implications of mixed factorial designs.

  • Address repeated-measures designs, identifying potential confounding variables and necessary controls.

Ethical Principles in the Study Context

  • Ethical Considerations:
      - Requires informed consent from adults and assent from minors, particularly concerning sensitive topics like fear induction.
      - Addressing the rights of control participants regarding treatment and potential psychological impact.

Summary

  • Factorial designs allow simultaneous exploration of multiple independent variables, revealing their interactions.

  • ANOVA provides robust tools for analyzing complexity within these designs, especially across mixed methodologies.

  • Important to consider ethical dimensions when involving vulnerable populations such as children in psychological research studies.