Abstract Within the task-based language teaching approach, the effect of input characteristics has been more centered on written than on aural texts. Moreover, the interplay of input characteristics and individual differences (IDs) has been insufficiently investigated. The current study sets out to address these gaps by comparing L2 learners’ comprehension of authentic and modified versions of two lectures and how IDs mediate the role of input condition in listening comprehension in an EAP context. The input was manipulated in terms of lexical sophistication and cohesion measures following previous research and informed by Natural Language Processing tools (e.g., Coh-Metrix, TAALES). Language aptitude, listening anxiety, task difficulty, task engagement, topic familiarity, and input condition (i.e., authentic and modified) were used as explanatory variables of listening comprehension as measured by 30 multiple-choice questions. The results of the Linear Mixed Effects Model suggested that the modified condition was significantly easier when IDs were controlled for, and engagement level and topic familiarity mediated the students' performance in the authentic and modified conditions. While high engagement and high topic familiarity were necessary for successful performance in the authentic condition, the opposite was true in the modified condition.

中文翻译：

---

输入特征和个体差异对真实和修改听力任务的二语理解的影响

摘要 在任务型语言教学方法中，输入特征的影响更多地集中在书面而非听觉文本上。此外，输入特征和个体差异 (ID) 的相互作用尚未得到充分研究。当前的研究旨在通过比较 L2 学习者对两个讲座的真实版本和修改版本的理解以及 ID 如何在 EAP 上下文中调节输入条件在听力理解中的作用来解决这些差距。根据先前的研究并通过自然语言处理工具（例如 Coh-Metrix、TAALES）提供信息，根据词汇复杂度和内聚措施对输入进行了处理。语言能力、听力焦虑、任务难度、任务参与度、话题熟悉度和输入条件（即，真实和修改）被用作听力理解的解释变量，通过 30 道选择题来衡量。线性混合效应模型的结果表明，当 ID 被控制时，修改后的条件明显更容易，并且参与水平和主题熟悉程度介导了学生在真实和修改后的条件下的表现。虽然高参与度和高主题熟悉度是真实条件下成功表演所必需的，但在修改后的条件下则相反。在真实和修改条件下的性能。虽然高参与度和高主题熟悉度是真实条件下成功表演所必需的，但在修改后的条件下则相反。在真实和修改条件下的性能。虽然高参与度和高主题熟悉度是真实条件下成功表演所必需的，但在修改后的条件下则相反。