Design Challenges in Pedagogical Agent Implementations
When creating pedagogical agents for use in online learning environments, designers face numerous challenges. These range from technological (e.g., How do I ensure proper lip-synching when speech is generated in real-time?) to pedagogical (e.g., How do I ensure that the agent provides scaffolding that is appropriate to the students’ needs at a given point in time?) to social (e.g., How can I develop an agent that is sensitive to students’ varying social needs?). While designers deal with these questions frequently and decide on what we deem to be the best approaches to tackle them, we don’t often share the our design thinking with others.
My colleagues and I (Gulz, Haake, Silvervarg, Sjoden, Veletsianos), have just published a book chapter that deals with this issue. In this chapter we discuss design challenges we faced when developing a pedagogical agent, and the steps we took, and decisions we made to tackle those challenges. The challenges we discuss are the following:
- how do we manage learners’ expectations of the agent’s knowledge and social profile,
- how do we deal with learners’ who engage in off-task conversations with an agent, and
- how do we manage abusive comments directed to the agent?
These issues were observed in studies that both Agneta Gulz and myself have independently conducted in the past, and sharing our design thinking with the community sounded like a great idea – hence the publication. A copy of this publication (1.7MB pdf) is provided below:
Gulz, A., Haake, M., Silvervarg, A., Sjoden, B., & Veletsianos, G. (2011). Building a Social Conversational Pedagogical Agent: Design Challenges and Methodological approaches. In Perez-Marin, D., & I. Pascual-Nieto (Eds.), Conversational Agents and Natural Language Interaction: Techniques and Effective Practices (pp. 128-155). IGI Global.
As always, I’d love to hear your input!