Gulsah Gurkan
Quantitative Behavioral Scientist | Behavioral Analytics & Insights
Welcome!
I’m an applied research scientist with a PhD in Applied Statistics and Measurement. I use causal inference, psychometric modeling, and multilevel analysis to understand human behavior at scale, and translate complex findings into clear, actionable insights.
Interests
- Causal inference and experimentation
- Psychometrics and measurement
- Behavioral and educational analytics
- Survey design and research
- Driving insights from large datasets
- Reproducible research
Education
PhD, Measurement, Evaluation, Statistics, and Assessment, 2021
Boston College, Chestnut Hill, MA, USA
MS & BS, Physics (minor in Education), 2011
Bogazici University, Istanbul, Turkey
Tools
R
Python
SQL
Projects
*
Contains an applied analysis of student retention using the Open University Learning Analytics Dataset (OULAD).
A technical case study using the public Telco Churn dataset (Kaggle) to demonstrate a rigorous causal inference pipeline.
A proof of concept using XGBoost to predict text complexity from keystroke logging dynamics.
Gulsah Gurkan, Yoav Benjamini, and Henry Braun. Published in Large-scale Assessment in Education, (2021).
An R package available on CRAN R repository providing a collection of psychometric methods for automated test assembly.
Henry Braun & Gulsah Gurkan. Published by: ETS Research Institute (2024).
From OLS to Multilevel Multidimensional Mixture IRT: A Model Refinement Approach to Investigating Patterns of Relationships in PISA 2012 Data.
Contains both a Python notebook and an R script for the same data cleaning and wrangling task to demonstrate the equivalent code structures.
Demonstrating how to create a heatmap using GeoJSON spatial data. Employing json files via web links is also exemplified.
R script that reads in and preps data files in an automated fashion, conditional on file type (e.g., .zip, .csv, .txt, etc.).
Code developed for practice sessions of a Multivariate Statistics course; topics covered such as logistic regression, principal component analysis, and discriminant analysis.