The Rutgers Artificial Intelligence and Data Science (RAD) Collaboratory

Biography

Professor Janice D. Gobert is a Cognitive Scientist. Her research and development work sits at the intersection of Learning Sciences, Science Education, and Computer Science, making both theoretical and applied contributions. Specifically, for the past two decades, Professor Gobert has been has addressing how computational techniques can be leveraged to support STEM teaching and learning aligned to reform policies such as the Next Generation Science Standards and 21st Century skill frameworks. This research has led to the development of a science learning and assessment environment called Inq-ITS (Inquiry Intelligent Tutoring System), which uses 3 (of Gobert et al.’s 6) patents on AI techniques, including knowledge-engineering, machine learning, and natural language processing to assess and scaffold students’ learning in real time at scale. That is, while middle and high school students conduct virtual science inquiry with Inq-ITS, AI algorithms automatically assess and scaffold students’ science competencies in real time; a pedagogical agent, Rex, driven by AI-algorithms jumps in to support students’ learning when the algorithms detects that the student is struggling. To support teachers, AI algorithms alert teachers in real time via a dashboard (Inq-Blotter) that identifies the students who need help and on what specific aspects (sub-components) of science inquiry practices they need help on. Additionally, Inq-Blotter provides empirically-tested prompts (TIPS) to guide teachers’ data-driven instruction to the whole class, to small groups (differentiated instruction), or to individual students.

With Inq-ITS technology, approximately 2 million science investigations have been completed by students across 50 states and 10 countries.

  • Inq-ITS’ algorithms can validly assess a wide range of science competencies and compare students’ competencies at “doing” science to their competencies at “writing” about science. This is critical for equitable assessment of language learners and those who may not be able to fully describe what they know in words,
  • Inq-ITS’ AI agent, Rex, demonstrably improves students’ competencies at science that transfer across topics and over long periods of time, and
  • Inq-Blotter, a teacher dashboard that runs alongside Inq-ITS, supports teachers’ real-time instruction of science via alerts and TIPS, which have been shown to lead to improvement of students’ science competencies.