Study Notes on Analysis of Variance

Introduction to Analysis of Variance

  • Instructor: Tony Jinx

  • Course Duration: Three weeks

  • Focus: Analysis of Variance (ANOVA) for Applications of Psychology Applied Research

  • Structure:

    • Six modules over three weeks

    • First three modules: Revision of basic ANOVA concepts and techniques (manual and SPSS)

    • Last three modules: Advanced topics in ANOVA

Overview of Experimental Research

Single Variable Experiments

  • Definition: Research studies examining only one independent variable

  • Simple Example: Two levels of an independent variable

    • Participants assigned to one of two groups based on the independent variable

    • Example Independent Variable: Liking for chocolate

    • Group 1: Low liking for chocolate

    • Group 2: High liking for chocolate

  • Control vs. Experimental Groups:

    • Control Group: Receives placebo or no treatment

    • Experimental Group: Receives actual treatment

    • Example:

    • Independent Variable: Memory treatment

    • Group 1: Receives drug (experimental)

    • Group 2: Receives placebo

Importance of Group Levels

  • Function: Determines if an independent variable has an effect

  • Types of Variable Levels:

    • Two levels because it's the only option available

    • Two levels due to research focus (e.g., comparing two memory techniques)

  • Single Variable Analysis Method: T-test for examining differences between two groups

    • Formula for T-test is provided but not required to memorize

Multi-Level Experiments

  • Definition: Experiments with more than two levels of an independent variable

  • Example: Alcohol's effect on driving performance

    • Initial two levels: 0 alcohol and low alcohol

    • Expanded to include: moderate alcohol and high alcohol

  • Insights gained from multi-levels:

    • A more complex understanding of alcohol's effects

    • Avoids interpolation errors by revealing non-linear effects

    • Example revelation: Alcohol effect on performance may not be linear

Ceiling and Floor Effects

  • Ceiling Effect:

    • Performance cap, e.g., 100% performance cannot be exceeded

    • Importance: Incorporating more levels needed to understand limits

  • Floor Effect:

    • Performance bottom limit, e.g., lowest performance possible

    • Visualization example provided

Analysis of Variance (ANOVA)

Understanding the Concept of ANOVA

  • Definition: Statistical method for analyzing multi-level data

  • Purpose: Examines differences among three or more levels of a single independent variable

  • Terminology: Independent variables are called factors in ANOVA

  • Relation to T-tests: ANOVA is akin to T-tests but for more than two groups

Conditions for Using ANOVA

  • Design Type: Appropriate when employing a between-groups design

    • Groups categorized by different levels of independent variable

    • True Experimental Design: Participants assigned to treatments

    • Natural Group Design: Participants divided based on characteristics

Comparison with Regression Analysis

  • Regression Analysis: Continuous measurement of independent variable

  • Example: Does age predict stress?

  • Comparison Methodology:

    • Regression: Measures individual scores directly

    • ANOVA: Categorizes into groups, e.g., young, middle-aged, old

    • Both methods analyze the same data but focus differently

Systematic vs. Unsystematic Variance

  • Systematic Variance (x): Variance attributed to the predicted model

  • Unsystematic Variance (y): Residual variance not explained by the model

  • Visualization: Diagram illustrating variance relationship is presented

SPSS and General Linear Model (GLM)

  • SPSS Role: Utilizes regression approach for ANOVA analysis

  • GLM Option: Allows examining regression slopes across categorical bars

    • Example Visualization:

    • Flat slope suggests no significant difference among means

    • Angled slope suggests significant difference

Traditional vs. Modern ANOVA

  • Traditional Hand-Calculated ANOVA: Compares variance among means directly

  • Variance Ratio Model: Derives understanding of analysis of variance process

  • Module Two Focus: One-way ANOVA and calculations by hand

Conclusion to Module One

  • Revision activity on fundamental concepts in ANOVA

    • Next topics will delve deeper into one-way ANOVA in module two