Week 11A Mediation(1) - Tagged

PY2501 Research Methods & Data Analysis

General Overview

• Course: Aston University, Birmingham, UK

• Focus: Mediation in research methods

• Presenter: Dr. Ryan D.Lhi

End of Module Feedback

• Links available for feedback surveys for courses PY4511 (MSc) and PY2501 (UG)

• Feedback period: Open for 2 weeks under "STUDENT VOICE" on Blackboard

Types of ANOVA

One-Way ANOVA

• Consider for >2 Levels

Repeated Measures ANOVA

• Focus on within-subject comparisons

Independent Groups ANOVA

• Between-subject comparisons

Decision Tree for Selecting Test Statistics

• Independent Groups ANOVA

  • Continuous DV, Nominal IV

• Repeated Measures ANOVA

  • Focus on paired samples

• Multiple Regression

  • Used when examining relationships between continuous variables

Mediation & Moderation Analysis

Extended Decision Tree

• Continuous DV, Continuous IV

• Determine if IV is mediated or moderated by another variable

• Mediation Analysis: assesses indirect effects

• Moderation Analysis: assesses interaction effects

Importance of Mediation and Moderation

• To clarify complex relationships between variables

  • Mediation: explains how one variable affects another

  • Moderation: explores conditions under which effects occur

Lecture Structure

• Part 1: Introduction to Mediation

• Part 2: Detailed exploration of Mediation

• Part 3: Application of Mediation in Jamovi

• Note: Moderation covered in Week 11B

Mediation: Official Diagrams

• Mediation Relationship:

  • Example: INCOME (IV) -> LIFE EXPECTANCY (DV)

  • Mediator: ACCESS TO HEALTHCARE

  • Total effect and direct effect explained

Example Relations in Mediation

Different Contexts

• Stove Example:

  • Stove Knob (IV) -> Water Temperature (DV), Mediator: Stove Temperature

• Video Game Example:

  • Storyline Quality (IV) -> Product Sales (DV), Mediator: Immersive Gameplay

• Influencer Example:

  • Content (IV) -> Brand Memberships (DV), Mediator: Trust in Brand

• Team Sports Example:

  • Crowd Support (IV) -> Team Victories (DV), Mediator: Player Motivation

Baron and Kenny (1986) Mediation Conditions

Four Necessary Conditions

1. IV significantly predicts DV

2. IV significantly predicts Mediator

3. Mediator significantly predicts DV

4. Direct effect of IV must be weaker than Total effect

• Implication on interpretations of mediation (complete vs. partial)

Practical Application in Jamovi

• Step-by-step use of mediation analysis

• Key paths in output:

  • Path a: IV to Mediator

  • Path b: Mediator to DV

  • Paths c' and c for direct and total effects

Summary of Mediation Effects

• Assessing conditions for mediation effects

• Evaluating whether mediation is complete or partial based on significance testing

• Note on quiz questions related to mediation

Conclusion

• Encouragement to utilize the email provided for questions.

• Email: r.blything@aston.ac.uk