A researcher's LUIDA workflow
The shape of a LUIDA study, from idea to data export.
A LUIDA study, whether it's a 5-minute pilot or a 500-participant run, tends to follow the same arc. Before heading into Setup or the tutorials, read this page once to get the shape in your head. While you're actually configuring and implementing your study, open the linked pages for any stage where you need the detail.
1. Decide what you want to measure
Before opening anything: write down your hypothesis, your independent and dependent variables, and roughly how many trials each participant will go through. LUIDA stays out of your stats plan — but its variable system is built around the within-subjects / between-subjects split, so it helps to know how you want to slice the design before you start clicking.
2. Register the study in the Web Console
Register the study's basics — title, consent form, eligible devices, room capacity, and any pre- and post-experiment questionnaires — on the Web Console. Keep the study in a "testing" status until you're ready for real participants.
Details: Components → Web Console
3. Build the experiment world in Unity
Open the LUIDA Unity Implementation Template and link it to the experiment you registered in step 2.
For simple designs, CCK and its LUIDA extensions are enough on their own. For slightly more complex designs, or when you want to keep CCK use to a minimum, enable LUIDA's experiment automation — declare your experiment variables, define a state machine for the trial flow, and assemble GameObjects that react to state changes (showing stimuli, recording responses, advancing trials).
Finally, upload the assembled world to Cluster (don't publish it yet) and register the world ID on the Web Console.
For the one-time install, see Setup next. After that, we recommend walking through Tutorial 0 — Hello LUIDA once to see the whole pipeline end-to-end before building your own. When you're implementing your own study and want the field-by-field reference for the Unity template, see Components → Unity Template.
4. Pilot it
In the world you uploaded to Cluster, run the experiment on yourself first, then have one or two colleagues try it. This is the stage for catching bugs and design issues.
5. Publish and collect
If nothing looks off, flip the study to published. Participants find it through LUIDA's Recruitment World and join from there; each time a session completes, the data is saved and shows up on the Web Console.
6. Download and analyze
Three datasets land in the Web Console: questionnaire answers, participant info, and any custom log data you recorded. Export as CSV or JSON, then continue in your usual stats workflow (R, Python, JASP, etc.).
Other
Paired or multi-participant studies, hardware integrations (EEG, mocap, eye tracking via OSC), and heavier pilot iteration all extend this arc rather than replace it. The Concepts and Tutorials sections cover the building blocks — and the docs site welcomes PRs for any gap you hit.