Conjoint experiments have become a popular tool in both social science and marketing research for understanding how people make decisions and what attributes matter most in their choices. Originally developed by marketers to measure consumer preferences for new products, conjoint analysis has expanded into political science, economics, sociology, and other fields as a powerful method for testing multiple causal effects simultaneously and eliciting respondent preferences. With a single well-designed survey, you can answer complex research questions about decision-making that would require dozens of traditional experiments.

In this course, you’ll learn the theory behind forced-choice conjoint experiments, how to design studies that answer your research questions, how to program and field these survey experiments in online platforms, and how to analyze and visualize the results using open source tools like R, Quarto, marginaleffects, and ggplot2. You’ll explore key causal estimands like average marginal component effects (AMCEs) and marginal means (MMs), as well as preference-focused measures like part-worth utilities and willingness-to-pay. By the end of the course, you’ll be able to design and field your own conjoint experiments, analyze the results with appropriate methods, and communicate your findings effectively with publication-ready tables and figures.

This course is designed to be hands-on and interactive. The first part of the course covers the fundamental principles and logic behind conjoint experiments and is more lecture-based, but in the remainder of the course, you will do practical activities. You’ll design a hypothetical experiment, build a conjoint survey using the open source R package surveydown, and analyze real-world experimental data with both frequentist and Bayesian methods to estimate two types of estimands:

• Causal effects, common in the social sciences, including marginal means (MMs), average marginal component effects (AMCEs), and average feature choice probabilities (AFCPs).
• Respondent preferences, common in marketing research, including part-worth utilities, willingness-to-pay, and simulated market shares.

You’ll leave the course with a complete workflow for conjoint analysis that you can adapt to your own research questions.