Built for Research
Better variable control, configurable AI, fluid human-AI switching, and structured conversation analytics. Stipends and free app use for qualifying studies.
Why Pangea Chat for Research
Better Variable Control
Both AI and human conversation partners on the same platform, giving better variable control than comparing tools across separate apps.
Configurable Implementation
We can work with researchers to control technical details of the AI, including making LLM responses deterministic for reproducible experiments.
Fluid Human ↔ AI Switching
Students fluidly switch between human and AI conversation partners. When given equal access to both, what do students prefer? Which produces better outcomes?
Richer Conversation Analysis
Vocabulary uses (correct/incorrect, with/without assistance), morphological concept usage, and interaction patterns. All structured data, not raw chat logs.
Open Core
The messaging client is open source on GitHub. Researchers can inspect data flows, audit privacy, and work with us to configure the AI tools for controlled experiments. Read our Privacy Policy.
Stipends Available by Application
Research stipends and free app use are available for qualifying studies. Mention your interest in the inquiry form below.
💡 Some Ideas
Research questions that Pangea Chat is uniquely well-suited to explore.
Human vs. AI Conversation Partner Comparison
Learners are assigned both human peer partners and the AI bot across a semester. The researcher compares vocabulary acquisition, grammatical accuracy gains, and engagement patterns between the two conditions, all on the same platform, with the same analytics, eliminating cross-tool confounds.
Effect of AI Writing Scaffolding on Interpersonal Communication Quality
Learners complete structured conversation activities with and without the AI grammar/vocabulary assistance tools enabled. The researcher measures whether AI scaffolding increases target-language use, reduces L1 fallback, and improves accuracy over time, using the platform's built-in vocabulary and grammar usage tracking.
Custom Course Development for an Underserved Language
An instructor of a less commonly taught language works with Pangea Chat to build a custom course of task-based conversation activities aligned to their curriculum. The study measures learner engagement, vocabulary growth, and proficiency gains across the semester using the platform's per-student analytics.
Peer Interaction Patterns in AI-Augmented Group Discussions
Learners participate in AI-moderated group discussion activities. The researcher analyzes who initiates, who responds, how the AI bot's prompts affect turn-taking and topic development, and whether learners choose to engage the bot or each other. Structured conversation data (messages, word-level feedback, morphological usage) is exported for analysis.
Interested in Researching with Pangea Chat?
Get in touch. We'll collaborate on study design, data access, and technical configuration to support your research goals.
