Paper Summary
Share...

Direct link:

Pedagogical Agents for Change: Renewing the Focus on Engagement

Sat, April 14, 2:15 to 3:45pm, Millennium Broadway New York Times Square, Floor: Sixth Floor, Room 6.01

Abstract

Research has demonstrated the critical role engagement plays in educational contexts. Learner engagement is associated with achievement and persistence in school as well as lower attrition rates. It also has been shown to be higher in learning environments that offer higher degrees of learner choice and classrooms with supportive teachers and peers. Various forms of engagement have been identified in the literature (e.g., behavioral, cognitive, emotional), with each contributing in important ways to learning and to subjective judgements of learning experiences. Although the topic of engagement has received considerable attention from educational technology researchers (e.g., educational games), the question of how best to design and deploy pedagogical agents in service of engagement has received much less attention.

So what changes when you add a pedagogical agent to a computer-based learning environment? Does the presence of an agent automatically provide a facilitating effect on engagement? The use of an agent is certainly a non-trivial design choice as it changes the experience to be (more of) a social situation. Activating a learner’s social frame in this way may produce an increase in engagement, at least initially; however, if a learner has previous experience with virtual agents or is hard to impress, there may be no such reaction. The challenge is to design pedagogical agents that, on top of having pedagogical and content knowledge, can deploy strategies to trigger and sustain learner engagement throughout a learning experience.

To identify and implement strategies for promoting engagement, some researchers have done what intelligent tutoring researchers have done for almost 40 years, which is to look at expert teachers. One finding coming from the literature on expert teaching is that engagement results from enthusiastic and passionate teachers. Inspired by this result, Lane et al. (2013) investigated the impact of enthusiasm by comparing two versions of a virtual human coach: one that was positive, energetic, and pro-active against one that was “all business” (i.e., task-focused, low-energy, and serious). The study occurred at a museum exhibit for computer programming of a robot, and involved composing small programs using a tangible interface. No differences were detected in terms of behaviors and task success (i.e., the information in both groups was the same), but that the group with the enthusiastic agent gave significantly higher self-efficacy ratings for computer programming. Some strategies used by the agent in this study included a higher degree of non-verbal signals of excitement (e.g., clapping, thumbs up), frequent praising (of effort), and various direct statements of joy (e.g., that it was fun to be there).

Much more work is needed to investigate the differential impact of varying strategies for inducing engagement. Because of increasing realism, conversational capabilities, and even the possibility to violate the laws of physics and go beyond what humans can do (e.g., a virtual character is not bound by gravity, can use magic, and so on), as well as the associated costs of implementing such strategies, it becomes very important to understand which design features produce the greatest impacts on engagement.

Author