This book collects the notes for BTE 693, Computational Methods and Field Experiments, a doctoral course on experimental design for business research at the University of Miami. Each chapter corresponds to one session of the course, in the order the material was taught, so the chapters that open with a “Summary of the Previous Session” are picking up exactly where the preceding chapter left off.
The book is organized in four parts.
- Part I – Foundations of Experimental Research.
-
Why experiments, and what secondary data cannot do; internal, external and ecological validity and the spectrum of experiment types; and the statistical machinery of hypothesis testing, power, and sample size.
- Part II – Optimizing Interventions.
-
The Multiphase Optimization Strategy (MOST), factorial and fractional factorial designs, and Sequential Multiple Assignment Randomized Trials (SMART) for interventions that adapt between stages.
- Part III – Adaptive and Real-Time Designs.
-
Just-in-Time Adaptive Interventions (JITAIs) and the Micro-Randomized Trials (MRTs) used to estimate their proximal effects.
- Part IV – Preferences, Scale, and New Tools.
-
Conjoint analysis for preference elicitation, the scalability problem and machine-learning estimation of heterogeneous treatment effects, and the use of large language models in designing and running experiments.
A single running example threads through most of the book: a smartphone stress-reduction app for university students, built from three components — mindfulness exercises, physical activity goals, and social support. The same app is used to illustrate a factorial MOST design, a two-stage SMART, a JITAI, a micro-randomized trial, and a choice-based conjoint study, which makes the differences between the designs easier to see than a fresh example for each method would.
Worked analyses are given in SAS. Code listings are numbered per chapter and indexed in the List of Listings at the front of the book. The data behind them are simulated, so the numbers are there to make the mechanics of each procedure concrete rather than to report findings.
Each chapter closes with the required readings assigned for that session of the course; the full references are collected in the bibliography. Appendix ? holds the readings for the course’s session on research ethics, which has no lecture notes of its own, together with the CITI training required before any human-subjects study at the University of Miami.