| This study aimed to explore the impact of pedagogical agents in computer-based listening instruction on EFL students' listening anxiety levels and listening comprehension skills. A total of 66 Korean college students received computer-based listening instruction. Students were randomly assigned to one of three conditions: American agent condition, Korean agent condition, or no-agent condition. Additional data sources were included in the experimental design in order to investigate students' learning experience more thoroughly.;Results indicated that there were no statistical differences in listening anxiety levels and listening comprehension skills between students who worked with the agent and students who worked without the agent. In addition, there was no statistical difference in listening anxiety levels between students who worked with the Korean agent and students who worked with the American agent. However, survey findings indicated a few differences between the agent condition and the no-agent condition when students were asked to describe their learning experiences. Students from both groups enjoyed the lesson overall; however, their comments revealed some differences. Students in the agent condition regarded the agent as an important contribution to their enjoyable learning experience, and specifically chose the presence of the agent as the reason they would want to work with the program again, while students in the no-agent condition mainly enjoyed the useful functions integrated into the computer-based lesson, and indicated they would work with the program again because it was interesting and helpful.;Although there were no statistical differences between the groups, these results seem to illuminate that the guidance provided by the pedagogical agent during the lesson positively affected students' learning experiences, which is in line with previous study findings. The findings from the survey can also provide suggestions regarding what aspects of pedagogical agents should be kept or improved for language learning. More data would strengthen the impact of the results. However, these findings offer practical and theoretical implications for using pedagogical agents in foreign language education. |
