If you're a psychology researcher, you may not have heard of oTree, but this software is a powerhouse of behavioral experiments and has received over 3,300 citations, mostly in behavioral and experimental economics. oTree is a versatile framework that lets you build highly dynamic experiments.
We are excited to announce that the oTree AI (OTAI) platform now supports psychology experiments! Not only the classics (Stroop, flanker, IAT, n-back, stop-signal, adaptive threshold estimation), but essentially any custom experiment that involves a stimulus-response, trial-based paradigm. This functionality is now available to psychology researchers for free.
There is no coding. You describe your desired experiment and an AI agent builds it for you.
Feel free to send this article to any psychology researchers you know!
See it for yourself
Our live demo site has ready-made versions of many classic paradigms that you can play in your browser right now (or import into your own OTAI project and customize):
- Attention and cognitive control: Stroop, flanker, Simon, go/no-go, stop-signal (SSRT), task switching, Posner spatial cueing, visual search
- Memory: n-back, old/new recognition
- Implicit attitudes: Implicit Association Test (IAT)
- Learning and decision making: two-armed bandit, adaptive quiz
- Perception and psychophysics: random dot motion with QUEST threshold estimation, audio perception tasks such as melody-in-noise detection with a staircase, and perceptual boundaries.
OTAI is highly trained in psychology experiments and can build games like these in minutes. The output is much more stable than what general-purpose AI coding tools produce.
Why use oTree instead of other frameworks?
So, why would you use oTree when there are already many software tools for psychology experiments, as well as AI coding tools like Claude Code and Codex?
Faster and more accurate than other AI coding tools
Our AI agent has much better accuracy than other AI coding tools. It uses the same newest LLMs, but is highly trained for building experiments and oTree specifically.
Everything built in
OTAI is cloud-based with no installation of any software; you build your experiment through AI chat (no coding).
The hosting and back-end are all built in and designed to work together. This includes all parts of a research study, including:
- Experiment hosting
- Integration with recruitment & payment platforms like Prolific, PayPal
- Consent forms
- Questionnaires
- Detailed live monitoring of participants and their responses
- Bots to automatically test your experiments
- Super-fast interactive testing of your experiments (you can test parts manually, then watch bots zoom through the other parts for you)
- Error monitoring: get alerts by email if any participants encounter errors.
- Chat between participants and experimenter
- Live experimenter input
This means you don't need to string together 5+ different pieces of software that were all created by different teams, or write custom scripts to fill in the gaps and missing functionalities.
Code that is readable and standardized
When you build your experiment with oTree, you get artifacts that are standard and consistent, and follow the same format & processes that thousands of other researchers are using. The code is easy for humans to read and audit.
This is in contrast to AI agents that "get the job done" but create a mess while doing so: complicated code, various workarounds, "glue scripts", and mixing in various technologies and different coding practices from what your peers are using.
Built for timing-sensitive research
Some highlights:
- Reaction times are measured from the moment the stimulus is actually painted on screen to the moment of the response, with precision on the order of 1 ms.
- Precisely timed trial sequences: fixation cross, stimulus, response window, feedback, inter-trial interval.
- Image and audio stimuli are preloaded so they appear without delay.
- Keyboard, button, click, and text responses.
- Practice blocks, multiple blocks, randomization, and trial-by-trial feedback.
- Built-in adaptive procedures (up/down staircases and QUEST).
- Responses can be evaluated in the browser, for the fastest possible pace between trials, or on the server, which keeps sensitive information like correct answers hidden from participants and lets you generate and queue trials dynamically, based on the participant's responses so far.
- Every response is saved as soon as it's made, so if a participant's browser crashes or reloads, they resume where they left off and no data is lost.
Get started
OTAI is currently free for psychology researchers. Just make an account at the below link then send an email to chris [at] otree [dot] org from your university email account with a link to your department website (or some other equivalent way for me to see your affiliation). People who have already used oTree or who are not primarily doing stimulus-response experiments are not eligible.