Tutorials

Analyzing Experiments using Causal Forests

This is a quick tutorial on how to analyze experiments using Causal Forests in in the {grf} R package. The tutorial features an analysis of an experiment on Rhea chicks that I got from Sue McDonnell who taught me a lot about science.
[slides][code][recap]

How to Test & Roll

This is a practitioners’ guide to my paper on Profit-Maximizing Marketing Experiments. Using R code and the replication files for the paper, it explains how to fit a metamodel to your own past A/B and then use that model to plan the sample sizes for future tests.
[slides & code on Github]

Advanced A/B Testing

After reviewing the basics of A/B test analysis, this tutorial covers heterogeneous treatment effects, uplift modeling, causal forests, pre-randomization blocking & matching, and post-stratification with example analysis in R.
[tutorial page] [github repo]

Is My Advertising Working?

This tutorial explains the differences between holdout experiments, model-based attribution, and media-mix modeling. The tutorial is built around analyzing single (synthetic) data set in different ways with data manipulation and analysis in R.
[slides & code on GitHub]

Choice Modeling for Marketing in R

Based on Chapter 13 of Chapman and Feit, these tutorials explain how conjoint analysis and choice modeling are used to inform product design. They also explain how to use the mlogit package in R.
[DataCamp course (fully-automated online course; requires subscription)] [slides]

Stan for Choice Modeling (with Kevin Van Horn)

This tutorial explains how to use the Stan platform to fit hierachical Bayes multinomial logit models in R. Assumes some experitise with MCMC output and choice modeling.
[slides & code on GitHub]

R for Marketing Research & Analytics (with Chris Chapman)

Chris and I developed a complete set of slides for our book, R for Marketing Research and Analtyics. Slides and code are provides for each chapter and follow the book closely.
[slides & code]

R for Reproducible Research

This tutorial for doctoral students in business explains why you should develop a reproducible workflow for your research and provides an overview of how to use R to automate and document your analysis. If you attend a live tutorial, you’ll get to see me do a live analysis of an A/B experiment on the growth of chicks (the baby bird).
[slides & code on GitHub]