Please join us online for a talk by Dr. ,听Professor of Political Science at the Universit茅 de Montr茅al. He will present 鈥�Model to Meaning: How to Interpret Statistical Models with R and Python鈥� on Wednesday, November 18, 2026, at 1:30pm Central Time. For access to this seminar, contact series administrator Cierra Streeter.
Vincent Arel-Bundock, PhD, is Professor of Political Science at the Universit茅 de Montr茅al, where he conducts research on political economy and research methods. He is an editor of the R Journal and the creator of several widely used software packages, including {marginaleffects}.听
Abstract:
Statistical and machine learning models are increasingly powerful and sophisticated. Unfortunately, the quantities that these models target are often difficult to interpret substantively, which makes it challenging for scientists to communicate their insights to colleagues and stakeholders. This seminar introduces a simple but powerful conceptual framework to translate complex model outputs into quantities that are substantively meaningful, intuitive, and accessible to a wide audience. The central idea is to abstract away from the mechanics of estimation and to treat models as "counterfactual prediction machines," which are subsequently queried to estimate quantities and conduct tests that matter substantively. Drawing on ideas developed in the book Model to Meaning, the seminar presents a model-agnostic workflow that can be applied in consistent fashion to draw inferences from a wide range of models, including models with categorical variables, interactions, nonlinearities, or hierarchical structures. It combines conceptual discussion with practical illustrations using the {marginaleffects} software package, which supports more than 100 different classes of models in R and Python. The goal is to provide researchers with a general framework and practical tools for interpreting statistical models.听