Alcohol, Music, Gender, and Oxytocin

Factorial ANOVA for Experimental Design

Statistics
Experimental Design
ANOVA
Factorial ANOVA studying alcohol, music, and oxytocin.
Published

June 15, 2025

Testing how alcohol, music, and gender affect oxytocin response

I analyzed a 3×2×2 factorial experiment to evaluate whether alcohol type, music genre, gender, or their interactions influenced blood oxytocin levels. The project focused on experimental design, three-way ANOVA, assumption checking, and interpretation of interaction effects.

View Source Code View Poster ← Back to Projects

Role
Statistical Analyst

Domain
Experimental Design

Stack
R · ANOVA · Diagnostics

Design
3×2×2 Factorial Experiment

Project Snapshot

Question

Do alcohol type, music genre, and gender influence oxytocin levels individually or through interaction effects?

Approach

Fit a three-way ANOVA model and evaluated main effects, two-way interactions, three-way interactions, and model assumptions.

Finding

Gender explained nearly all variation in oxytocin response, while alcohol type and music genre showed no meaningful main or interaction effects.

Overview

Oxytocin is often associated with social bonding and emotional response, making it an interesting outcome for studying how external stimuli may affect hormone levels.

This project tested whether alcohol type, music genre, and gender were associated with differences in oxytocin response using a controlled factorial experiment. The main goal was not only to test individual factor effects, but also to determine whether combinations of factors produced interaction effects.

The Challenge

Multiple experimental factors

The study included three predictors: alcohol type, music genre, and gender, creating a full 3×2×2 factorial design.

Interaction effects

The analysis needed to test whether the effect of one factor depended on the level of another factor.

Model assumptions

ANOVA requires checking assumptions such as normality, constant variance, and independent errors.

Clear interpretation

The final results needed to explain statistical significance in practical terms, not just report p-values.

Analysis Pipeline

1
🧪

Experimental Design

Defined a 3×2×2 factorial structure with alcohol type, music genre, and gender.

2
📊

Data Exploration

Summarized oxytocin levels across factor combinations.

3
📈

Three-Way ANOVA

Estimated main effects, two-way interactions, and the three-way interaction.

4
🧾

Assumption Checks

Evaluated residual normality, variance stability, and diagnostic plots.

5
🔍

Effect Interpretation

Compared effect sizes and statistical significance across factors.

6

Conclusion

Translated ANOVA results into clear experimental findings.

End-to-end statistical workflow for a factorial experimental design.

Technical Stack

  • R
  • ANOVA
  • Factorial Design
  • Three-Way ANOVA
  • Interaction Effects
  • Residual Diagnostics
  • Hypothesis Testing
  • Data Visualization
  • Statistical Inference

What I Built

Factorial Design Analysis

Structured the experiment as a 3×2×2 factorial design to test main effects and interaction effects simultaneously.

Three-Way ANOVA Model

Fit an ANOVA model to estimate the influence of alcohol type, music genre, gender, and their interactions on oxytocin levels.

Assumption Diagnostics

Checked model assumptions using residual analysis and diagnostic plots to validate the reliability of the ANOVA results.

Results Interpretation

Translated statistical output into a clear explanation of which factors meaningfully affected oxytocin response.

Key Results

🧬
~99%

Variance explained by gender

🧪
3×2×2

Full factorial experimental design

📊
ANOVA

Main effects and interactions tested

What I Learned

This project strengthened my understanding of how factorial experiments are analyzed and interpreted. The most important takeaway was that interaction effects are central to experimental design because they reveal whether one factor changes the effect of another.

I also learned the importance of validating model assumptions before interpreting statistical results. Even when an ANOVA table produces significant p-values, residual diagnostics and practical interpretation are essential for drawing reliable conclusions.

Research Deliverable

The full research poster includes the experimental setup, ANOVA results, diagnostic checks, and final interpretation.

View Source Code View Poster (PDF) ← Back to Projects