eColloq Webinar: Implicit Learning Assessments, and Research on Pedagogical Agents

Tuesday October 24, 2017 from 12-1:30 pm Pacific / 3-4:30 pm Eastern

Register for this webinar

The first webinar in a new eColloq Series on Cyberlearning will feature presentations by Jodi Asbell-Clarke and H. Chad Lane on their research, followed by discussion. The eColloq will be chaired by Michael Hoffmann, Georgia Institute of Technology.

We Know More Than We Can Tell: Differentiation through Implicit Learning Assessments

Jodi Asbell-Clarke, TERC

Schools test for what learners can express in text and other written or verbal expressions, but this may limit our ability to measure what learners really know. Research shows that some learners with cognitive differences such as ADD, Dyslexia, and Autism may be exceptionally skilled at certain types of problem-solving involving idea generation and systematic thinking. Researchers and educators need methods to assess what learners know, their implicit knowledge, not just what they can tell. Educational Data Mining (EDM) in digital learning environments such as digital games allows this kind of implicit learning assessment. I will present examples of how implicit learning assessments use EDM to study STEM learning in a physics game and a computational thinking game, how this information can be used for differentiated learning opportunities, and the important role of teachers in bridging implicit to explicit classroom learning.

Research on Pedagogical Agents: How Making Computers More Human-like Can Improve Learning

H. Chad Lane, University of Illinois Urbana-Champaign

From the earliest days of intelligent tutoring systems, which tended to be impersonal and “all business”, researchers have pursued more human-like and engaging forms of intelligent support for learning. Through natural language interaction, support for emotional aspects of learning, animation, sound and more, the goal of increasing the communicative bandwidth of machines and emphasis on more human-like interactions has brought radical change to what it means to learn from a machine. In many ways, the challenges associated with these advances come together in the design and deployment of animated pedagogical agents (PAs), which are virtual characters embedded within digital learning environments that seek to enhance learning. In this talk I will review the Cyberlearning community’s progress on the challenge of developing PAs, presenting several examples that illustrate the many roles that PAs can play in support of learning (including teacher, coach, peer, role-player, or adversary). I will summarize lessons learned from my experiences deploying PAs in informal learning environments, and address the many risks of haphazardly including a PA in a learning environment. Finally, I will review key empirical findings that PAs have enabled researchers to pursue and conclude with a discussion of different visions of the future of PAs, including thoughts on associated ethical and practical aspects of their adoption.

Suggested Citation

Asbell-Clarke, J. (2017, October 24). We Know More Than We Can Tell: Differentiation through Implicit Learning Assessments. Retrieved from

Lane, H.C. (2017, October 24). Research on Pedagogical Agents: How Making Computers More Human-like Can Improve Learning. Retrieved from

Share this: