The online human behaviour lab of the future
PsyNet is a powerful new Python package for designing the next generation of online behavioural experiments. It streamlines the development of highly complex experiment paradigms, ranging from simulated cultural evolution to perceptual prior estimation to adaptive psychophysical experiments. Once an experiment is implemented, it can be deployed with a single terminal command, which looks after server provisioning, participant recruitment, data-quality monitoring, and participant payment. Researchers using PsyNet can enjoy a paradigm shift in productivity, running many high-powered variants of the same experiment in the time it would ordinarily take to run an experiment once.
PsyNet architecture. PsyNet integrates with the Dallinger platform for orchestrating server provisioning and participant recruitment via Amazon Mechanical Turk. Participants take part via their web browser, which interacts with a server cluster hosted on Heroku.
PsyNet dashboard. PsyNet provides an advanced dashboard for monitoring all aspects of the experiment lifecycle, including the graph structure of the database (right side of the figure), participant recruitment (left side of the figure), and payment.