Educators use various strategies to increase listening comprehension for nonnative English speakers in the classroom and multimedia environments. Research on audio reduction has shown mixed results, whereas a study that enhanced (doubled) virtual human gesturing found increased listening comprehension with procedural information (Davis and Vincent, British Journal of Educational Technology 50:3252–3263, 2019). This research examined the use of virtual human gesture frequency (enhanced, average, no) and video speed (normal, reduced 25%) on participant satisfaction, perception of agent persona, and learning outcomes with procedural information. Structural equation modeling was used to analyze the data from 234 multinational university students. The results indicated that normal video speed significantly increased satisfaction compared to reduced speed; agent persona was rated significantly higher with agents that gestured compared with the no-gesture condition; and enhancing the gesture frequency significantly increased learning outcomes compared to the average and no-gesture conditions. These findings support previous studies that indicated enhanced gestures significantly increase the learning of procedural information. Also, agent gesturing increased the perception of the agent as having more persona, which supports systematic review findings that gesturing is a principal moderator for agent persona (Davis et al., Journal of Research on Technology in Education 53:89–106, 2021). However, this research provides evidence that a 25% reduction in video speed might be too slow to maintain satisfaction with advanced foreign language users and that less reduced rates such as 15% or 10% need to be considered. Finally, this research puts forth a gesture design framework for designers to create gesturing virtual humans in multimedia environments.

教育工作者在课堂和多媒体环境中使用各种策略来提高非英语母语者的听力理解能力。对音频减少的研究显示出不同的结果，而一项增强（加倍）虚拟人类手势的研究发现，通过程序信息提高了听力理解能力（Davis 和 Vincent，British Journal of Educational Technology 50：3252–3263，2019）。该研究检查了虚拟人类手势频率（增强、平均、否）和视频速度（正常，降低 25%）对参与者满意度、代理角色感知和程序信息学习结果的使用。结构方程模型被用来分析来自 234 名跨国大学生的数据。结果表明，与降低速度相比，正常视频速度显着提高了满意度；与不做手势的情况相比，有手势的代理的代理角色评分明显更高；与平均和无手势条件相比，提高手势频率显着提高了学习成果。这些发现支持先前的研究，即增强的手势显着增加了程序信息的学习。此外，代理手势增加了代理具有更多角色的看法，这支持系统审查结果，即手势是代理角色的主要调节器（戴维斯等人，教育技术研究杂志 53：89-106，2021 年） . 然而，这项研究提供的证据表明，视频速度降低 25% 可能太慢而无法保持高级外语用户的满意度，需要考虑降低 15% 或 10% 等较少的降低率。最后，本研究提出了一个手势设计框架，供设计师在多媒体环境中创建手势虚拟人。