🚨 Heads up on upfront costs: This course requires a Claude Code subscription and a budget for Prolific participants (Which does not have to be much depending on your study size). Plan for these before you begin. The payoff: you'll go from zero to a fully deployed, analyzed experiment end to end in a single day.
Intro to the Course
Overview
This course is about a revelation. At some point in every researcher's career, something clicks. You stop fighting your tools and start building them. This course is about that moment — and giving it to you on purpose. We'll walk you through how to design, deploy, and analyze behavioral experiments using a modern full-stack toolkit. No more wrestling with survey platforms that weren't built for science. No more duct-taping together tools that almost do what you need. You're going to build exactly what your research demands.
The A-Ha Moment
There's a feeling that hits when you deploy your first custom experiment to the internet. You built it. You own it. You can collect behavioral data from anywhere in the world, on your terms, with full control over every pixel and every data point. End to end — in a single day. Compare that to the alternative: configuring a platform that wasn't designed for experiments, bending it into shapes it wasn't meant to hold, and still ending up with data that's hard to work with. This course is the alternative.
What We're Building
By the end of this course, you will have:
- Designed and deployed a real behavioral experiment to the internet
- Collected data from real participants using Prolific
- Analyzed that data using Python's scientific stack
- Used AI tooling to accelerate every step of the process
This isn't a survey of tools. It's a complete workflow — from research question to analyzed dataset, finished in a day.
Who This Is For
Behavioral data scientists who are ready to meet the tools of the future. If you've ever thought there has to be a better way while clicking through a form builder or exporting CSVs from a platform that charges you per response — this course is for you. No prior web development experience required. Curiosity required.
The Stack
💡 We'll be building with a specific stack throughout this course, but the concepts — experiment design, participant recruitment, data collection, and AI-assisted analysis — generalize to other frameworks. The stack is a vehicle. The thinking is what transfers.
Experimentation
| Tool | Purpose |
|---|---|
| Next.js | An opinionated framework for building the experiment interface |
| PostgreSQL | Relational database for storing participant data |
| Prisma | Type-safe database Object Relational Mapper |
| Tailwind CSS | Utility-first styling |
| TypeScript | Type-safe JavaScript |
| Prolific | Participant recruitment |
| Supabase | Managed database |
| GitHub | Version control and deployment |
Data Science
| Tool | Purpose |
|---|---|
| Python | Primary analysis language |
| Pandas | Data wrangling and manipulation |
| Scikit-learn | Statistical modeling and Machine Learning |
| NumPy | Numerical computing |
| Matplotlib | Visualization |
| Seaborn | Statistical data visualization |
| Tensorflow | Machine Learning |
AI Tooling
| Tool | Purpose |
|---|---|
| Claude Code | AI-assisted development throughout the entire workflow |
Why Now?
The ability to scale software up and down has never been easier. The ability to harness behavioral data has never been more accessible. The tools are too good not to share.
Modern infrastructure — cloud databases, serverless deployments, managed auth, AI-assisted coding — has collapsed what used to take a team weeks into something one researcher can build in a weekend. The barrier between I have a research question and I have data has never been lower.
Course Structure
Each lesson follows the same pattern:
- Concept — Why this matters for behavioral research
- Code — Working implementation you can run
- Key Points — What to take away
- Exercises — Hands-on practice
We move fast. We build real things. We don't stop until you have a running experiment.
Your Instructors
Dr. David J. Cox — Dr. David Cox can formally lay claim to being a bioethicist (master's degree), a board-certified behavior analyst at the doctoral level (PhD degree), a behavioral economist (post-doc training), and a data scientist (post-doc training). He has worked in behavior analysis for 20 years as a clinician, academic researcher, scholar, technologist, and all-around behavior science junky. From his work and collaborations, David has published 70+ peer-reviewed articles, book chapters, and books. And, has had the fortune to serve as Editor in Chief for The Experimental Analysis of Human Behavior Bulletin and Associate or Guest Editor for Perspectives on Behavior Science, Behavior Analysis in Practice, Journal of Applied Behavior Analysis, Psychological Record, Education and Treatment of Children, Toward Data Science, and Behavior and Social Issues. When he's not doing research or building quantitative models of behavior-environment relations, he enjoys spending time with his wife, two beagles, and two kittens around St. John's, FL.
Jacob Sosine — Jacob began his work in behavior analysis in 2011, directly implementing behavior analytic services. He holds a Bachelor's degree in Exercise Physiology from Chico State University, a Master's in Professional Behavior Analysis from Florida Institute of Technology, and an MBA with an emphasis in Business Analytics from Saint Mary's College of California. As a Board Certified Behavior Analyst and data scientist, Jacob's research sits at the intersection of behavior analysis, technology, and machine learning/artificial intelligence. Jacob is the founder of Behaviorchain LLC, where he builds technology for behavior analysts — tools designed to make research more accessible, data more actionable, and the field's growing literature easier to navigate. His work focuses on developing innovative software and data-driven solutions that empower behavior analysts in both clinical and research settings. Outside of work, Jacob is convinced the S.F. Giants will win the World Series every year, and you can often find him at a game with his wife and young daughter. He is also on record as claiming that Dr. Cox shares his belief that the SF Giants are the most storied franchise in all of sports — a claim Dr. Cox has neither confirmed nor denied.
This course will provide continuing education units (CEUs). Details will be announced ahead of the full launch.
Coming Summer 2026
🗓️ This course launches Summer 2026. Want to be notified when it drops? Express your interest here and you'll be first to know.
Key Points
- Modern tooling changes what's possible — You can now build, deploy, and analyze a behavioral experiment in days, not weeks
- Ownership beats configuration — A custom-built experiment gives you full control over data, design, and participant experience
- The stack is learnable — Each tool in this course was chosen because it's powerful and approachable
- AI accelerates everything — Claude Code isn't a shortcut; it's a force multiplier for researchers who know what they want to build