Type of Document	Dissertation
Author	Kim, Yanghee
Author's Email Address	yhkim03@hotmail.com
URN	etd-07092004-133044
Title	Pedagogical Agents as Learning Companions: The Effects of Agent Affect and Gender on Learning, Interest, Self-Efficacy, and Agent Persona
Degree	PhD
Department	Educational Psychology and Learning Systems, Department of
Advisory Committee	Advisor Name Title Amy L. Baylor Committee Chair Ian Douglas Committee Member John M. Keller Committee Member Marcy P. Driscoll Committee Member
Keywords	pedagogical agents learning companions educational technology computer-based learning Affective Pedagogical agents
Date of Defense	2004-06-29
Availability	unrestricted
Abstract The purpose of the study was to examine the impact of the affect and gender of pedagogical agents as learning companions (PALs) on learning, interest, self-efficacy, and agent persona. Two experiments were conducted to examine PALs’ affect separately in terms of affective expression and affective response. 142 students in a computer literacy course participated in Experiment I,which examined the effects of PALs’ affective expression (positive vs. negative vs. neutral) and gender (male vs. female). 56 pre-service teachers participated in Experiment II, which examined the effects of PALs’ affective response(responsive vs. non-responsive) and gender (male vs. female). The results of Experiment I indicated that affective expression significantly influenced agent persona: Students who worked with the PAL expressing positive affect perceived the PAL’s persona more positively than students who worked with the negative PAL (p < .05). Also, students who worked with the PAL who did not express affect (neutral) perceived the PAL’s persona more positively than students who worked with the negative PAL (p < .05). Next, PALs’ gender significantly affected learning, interest, and agent persona: Students who worked with a male PAL learned more (p < .01), were more interested in the task and the PAL (p < .05), and tended to perceive the PAL’s persona more positively than students who worked with a female PAL (p = 0.7). Also, there was a significant interaction effect between affective expression and gender on agent persona (p < .05): When the PALs expressed positive affect, the persona of a male PAL was perceived more positively than that of a female PAL; however, this interaction was minimal when the PALs expressed negative or neutral affect. This interaction trend was also consistent for learner interest. The results of Experiment II revealed that PALs’ affective response significantly influenced interest and self-efficacy: Students who worked with a PAL that responded to their affect showed significantly higher interest (p < .05) and self-efficacy beliefs (p < .05) than students who worked with a PAL that did not respond to their affect. Next, PAL gender significantly influenced learners’ perceptions of PALs’ persona and partially learner interest: Students perceived the persona of the male PAL more favorably (p < .05) and showed higher interest when they worked with the male PAL (p = .07). Overall, PALs’ affect and gender influenced learners’ affective and cognitive characteristics as real human teachers or peers did (Wong & Dornbusch, 2000). This implies that PAL-learner relationships in computer-based learning can be consistent with human peer relationships in traditional classrooms. Given the finding of the study confirming the instructional impact of PALs’ affect and gender, researchers can design the gender and affect of PALs appropriately in the way that facilitate learning and motivation in computer-based environments.
Files	Filename       Size       Approximate Download Time (Hours:Minutes:Seconds)     28.8 Modem   56K Modem   ISDN (64 Kb)   ISDN (128 Kb)   Higher-speed Access    Yanghee_Dissertation_v2.pdf 1.23 Mb 00:05:41 00:02:55 00:02:33 00:01:16 00:00:06