In social science research, asking “is there a relationship?” is often not enough. Researchers now seek deeper answers like “how does this relationship occur?” and “under what conditions is it stronger or weaker?” This is where mediation and moderation analyses come into play. In this article, we’ll explain these two powerful analysis types in simple terms.

1. What Is Mediation Analysis?

Mediation analysis tests whether the effect of an independent variable on a dependent variable occurs through a third variable (the mediator).

Example:
Study time → (via self-confidence) → exam success

In other words, study time may directly affect success, but it may also increase self-confidence, which in turn affects success indirectly.

Model Structure:

X → M → Y
X: Independent variable
M: Mediator
Y: Dependent variable

Analysis Methods:

• Baron & Kenny method
• Sobel test
• Bootstrap method (most recommended)

2. What Is Moderation Analysis?

Moderation analysis tests whether the relationship between an independent and a dependent variable is strengthened or weakened by a third variable.

Example:
Study time → exam success
This relationship may vary depending on the student’s motivation level.

Model Structure:

X × Z → Y
X: Independent variable
Z: Moderator
Y: Dependent variable

Tip:
Add an interaction term (X × Z) to the model to test moderation.

3. Differences Between Mediation and Moderation

4. How to Run the Analysis in SPSS with PROCESS

PROCESS Macro (by Andrew F. Hayes) is an SPSS add-on that allows easy mediation and moderation analysis.

• Model 4: Mediation analysis
• Model 1: Moderation analysis
• Models 7, 14, 58: Combined models (mediation + moderation)

Steps:

1. Install the PROCESS macro
2. Go to Analyze > Regression > PROCESS
3. Enter the model number
4. Define the variables
5. Set bootstrap sample size (e.g., 5000)
6. Click “OK”

5. How to Report in Your Thesis

“Mediation analysis was conducted to examine the role of self-confidence in the effect of study time on exam success. The indirect effect was found to be significant using the bootstrap method (Indirect Effect = 0.42, 95% CI [0.18, 0.67]).”

“Moderation analysis was conducted to test whether motivation level moderates the relationship between study time and exam success. The interaction term was significant (β = 0.31, p < 0.01), indicating that motivation strengthens the relationship.”

6. Conclusion

Mediation and moderation analyses not only identify relationships in social science research but also reveal their mechanisms and conditions. Including these analyses in your thesis enhances both the depth and academic value of your research.