Designing a language specific data collection platform with gamified results and open data

Erin M. Buchanan

Outline

• Linguistic Data Collection
• ManyLanguages Community
• The Platform
• Test Case

The World’s Languages

Science’s Languages

ManyLanguages

• A big team science community supported by the Psychological Science Accelerator
• Our mission is to facilitate the connection between language science researchers to diversify the languages, participants, researchers, and projects represented in the language sciences.

ManyLanguages

• https://chenchenzi.github.io/manytones/
• Pilot data collection / registered report stage

The Platform

• Open architecture / infrastructure for data collection and sharing
• Dockerized platform that connects the participants, researchers, data collection, and sharing components

The Platform: Users

• Participants:
  • Sign up and complete the studies
  • Review the results / feedback on your results
• Researchers:
  • Launch studies
  • Review results / download the data
  • Export anonymized datasets

The Platform: Users

• Administrators:
  • Review uploaded studies for adherence to platform requirements
  • Ensure data sharing anonymity
  • Assist in feedback programming

The Platform: Tech Specs

The Platform: Tech Specs

The Platform: Tech Specs

The Platform: Example Screens

The Platform: Example Screens

The Platform: Example Screens

The Platform: Example Screens

The Platform: Sustainability

• The platform is “pay to close” - it’s free to post for researchers who will use JATOS + share data
• Researchers can use other software and/or not share data by paying a small fee to post the study

The Platform: Test Case

The Platform: Test Case

• 45 languages across the globe including multiple writing systems
• Collect priming data to “finish” languages based on previous power estimates
• Collective subjective data: age of acquisition, familiarity, valence, arousal, concreteness, and imageability

Final Thoughts

• Linguistics is ripe for studies in this area to increase the breadth of language, researcher, and participants included
• Digital infrastructure can advance open, global, and interdisciplinary science
• Citizen science to engage the public given the increase in interest in language due to AI / LLMS

Final Thoughts

• Go visit Addie’s poster to learn more!
• Join us for the study:
  • Authorship on three research papers as part of dissertation
  • Other side projects expected
  • Funding is available for low resource languages
• Should start in 2026, continue data collection throughout the year based on translation / lab timelines