Code & Resources
Tools I build and reuse across projects — survey instruments, replication code, and agent skills that automate the repetitive parts of empirical research. Everything here is free to use; if something breaks or you want to adapt it, email me.
Skills
Packaged instructions that give an AI coding agent (Claude Code, and similar tools) a reliable, repeatable procedure for a research task — so the work is done the same way every time instead of being re-improvised in each conversation.
QSF Builder
Writes Qualtrics Survey Format (.qsf) files — the JSON that Qualtrics accepts under Survey → Import Survey. The format is undocumented and unforgiving: several fields have to agree with each other, and one mismatch makes the import fail without saying why. The skill carries the structure the agent needs, so you upload a single file instead of adding questions and pasting JavaScript one at a time in the editor.
It covers question types and validation, survey flow and branching, embedded data, display and skip logic, question JavaScript (including calls to the OpenAI API from inside a survey), and editing an existing .qsf without disturbing the rest of it.
Use it when: building a Qualtrics survey that would be slow to assemble by hand — many questions, custom JavaScript, embedded data, display logic, or an API call from inside the survey. Input → output: a description of the survey, or an existing .qsf plus the changes you want → a complete .qsf ready to import.
Packages
First entries coming shortly.
Research Resources
Survey instruments and companion material from my papers, packaged so other researchers can reuse them.
AI-Led Interview Toolkit
An interview led by a large language model, run inside an ordinary Qualtrics survey — the module from Is Money Overrated?, packaged so it can be dropped into someone else’s survey. The respondent describes a real decision in their own words, the model asks up to three adaptive follow-up questions, and it then generates a personalized binary choice you can use in ordinary closed-ended outcome questions.
Includes an importable .qsf, an implementation guide, and the four interview prompts. Basic use requires no programming.
Replication Material
Code and data for my papers are posted alongside each paper. See the Research page for links — including the online appendix and pre-registration for Is Money Overrated?
Toolkit
Stata · Python · R · Git · LaTeX