The problem
Psychophysics research often needs a participant to follow a moving target while the software records what they do and when. Off-the-shelf tools are either too rigid for a custom experiment or too expensive for a student lab.
What it does
TrackingExperiments is a desktop prototype that presents a moving object and captures the participant’s tracking behaviour. It was built to be adapted quickly: the experiment loop, visual presentation, and data capture live in one small application instead of being spread across a stack of scripts.
How it’s built
The application is written in C# on .NET. Keeping it as a single desktop program made it possible to run on lab machines without a server, and to reason about timing in one place.
What I took away
- Real-time interaction is mostly a timing and state problem, not a graphics problem.
- Research software has to be boring in the right places: reproducible runs, predictable output, no surprises.
- The hardest part of a prototype is deciding which knobs to expose and which to hard-code.
Next steps
Packaging the configuration so a researcher can define a new experiment without touching the source, and adding a cleaner export format for analysis.