$\ $

Item Response Theory and Test Construction in Python

$\ $

This repository contains the psy126 seminar materials for the Neurocognitive Psychology M.Sc. at the University of Oldenburg. All lectures and exercises are provided as a Jupyter Book and focus on applying Item Response Theory (IRT) and test construction techniques in Python.

Topics covered

The book includes notebooks and slides on:

• Setting up your Python environment and working with Jupyter notebooks
• Data curation, manipulation and visualization with pandas, numpy and matplotlib
• Measurement models for dichotomous items (Rasch, 2‑PL, 3‑PL) using rpy2
• Models for polytomous items: Rating Scale Model, Partial Credit Model, Generalized PCM and Graded Response Model
• Concepts of quantitative IRT, differential item functioning, and multilevel / multidimensional modeling
• Templates for test theory reports and further resources

Online usage (recommended)

The book is best viewed online via GitHub Pages:

Local usage

To build the book locally:

cd <path/to/book/>
pip install -r requirements.txt
jb build .

This creates HTML files in _build/. Open _build/html/index.html in your browser to browse the book. Exercise notebooks live in the book/ folder and can be run locally or on Google Colab.

Cloning the repository

Clone the repository if you want to keep a local copy and pull updates:

git clone https://github.com/leonardozaggia/psy126.git

Save your exercise solutions outside the repository to avoid overwriting them when running git pull.

A comprehensive tutorial on using Git in VS Code is available here: