Twin2K: What the Data Shows
Overconfidence is positive across every demographic slice. An interactive dashboard exploring behavioral economics across 2,058 U.S. adults.
We Built the Most Sophisticated Personality Model We Could. A 3-Variable Linear Model Still Won.
A single incentivized trust game score outperformed all five Big Five personality scales in predicting financial risk tolerance.
Twin2K: What the Data Shows
An interactive dashboard built from the flattened Twin2K parquet file — demographics, personality, cognition, wellbeing, and behavioral economics in one place.
We Built the Most Sophisticated Personality Model We Could. A 3-Variable Linear Model Still Won.
Applying the full CRISP-DM pipeline to behavioral psychometric data — and what the results say about the limits of personality-based prediction.
We Built the Most Sophisticated Personality Model We Could. A 3-Variable Linear Model Still Won.
A full CRISP-DM data mining pipeline applied to the Twin-2K-500 dataset, from business problem framing through model validation.
Twin2K: What the Data Shows
Five tabs, 22 interactive charts, one dataset. Explore the Twin2K behavioral survey across demographics, personality, cognition, wellbeing, and behavioral economics.
We Built the Most Sophisticated Personality Model We Could. A 3-Variable Linear Model Still Won.
Three independent modeling methods converged on the same answer — what that convergence reveals about prediction limits in behavioral data.
Research Organization Templates for Students and Analysts
Two free templates for organizing academic articles and data sets—because research without organization is just reading.
We Built the Most Sophisticated Personality Model We Could. A 3-Variable Linear Model Still Won.
What the gap between industrial digital twins and human behavioral models reveals about the limits of psychometric prediction.
We Built the Most Sophisticated Personality Model We Could. A 3-Variable Linear Model Still Won.
The gap between academic behavioral measurement and deployment-ready accuracy in robo-advisors, insurance pricing, and fintech behavioral products.
We Built the Most Sophisticated Personality Model We Could. A 3-Variable Linear Model Still Won.
A Bootstrap Forest detected personality signal that linear models missed — but still couldn't beat a 3-variable regression on out-of-sample prediction.
We Built the Most Sophisticated Personality Model We Could. A 3-Variable Linear Model Still Won.
Project work from TBANLT 485 (Business Intelligence) at UW Tacoma, applying CRISP-DM to the Twin-2K-500 behavioral dataset.
Research Organization Templates for Students and Analysts
Digital note cards for the modern researcher—systematic approaches to literature reviews and data exploration.
Research Organization Templates for Students and Analysts
Structured frameworks for bachelor's and graduate work in research-oriented and STEM fields.
Research Organization Templates for Students and Analysts
Downloadable Word templates for deep analysis (30 min) and quick-scan (10 min) research organization.