Unsupervised Machine Learning for Social Cognition

Using unsupervised machine learning to explore patterns in beliefs and attitudes.

A tutorial and applied methods paper where I show how K-means, DBSCAN, PCA, and market basket analysis can reveal structure in large-scale attitude data.

Open paper Open PDF Back to projects

Case study

Inside this project.

A concise view of the question, design, my role, method, contribution, and scope.

Question: How can social-cognition researchers explore structure in attitude data without imposing predefined categories?
Data / design: Methods tutorial using Project Implicit Race IAT data and worked examples.
My role: Lead author.
Method / approach: K-means, DBSCAN, PCA, and market basket analysis, used as discovery tools that still need theory-based interpretation.
What it contributes: A tutorial and applied methods paper where I show how K-means, DBSCAN, PCA, and market basket analysis can reveal structure in large-scale attitude data.
Venue / status: Published, Social Cognition, 2025. Open paper
Scope: The clusters and dimensions are exploratory patterns that require theory-based interpretation.

Study details

Project Implicit Race IAT data
K-means, DBSCAN, PCA, and market basket analysis
Theory check between model output and social-cognition constructs
Published in Social Cognition