Abstract Background Due to MOOC high attrition rates, this study aims to assess differences in self-regulated learning strategies and other variables related to MOOC retention (perceived effectiveness, MOOC interaction, motivation and socio-demographic characteristics) between course completers and non-completers. This work also aims to translate into Spanish and validate an instrument for the assessment of self-regulated learning (SRL) behaviours in MOOCs. Materials and methods: 582 participants answered the translated SRL questionnaire and other questions related to MOOC retention. The comparison between MOOC completers and non-completers was carried out in a subgroup of 176 undergraduate students. Results: Completer students were more capable of self-regulating their learning and showed significantly higher levels of perceived effectiveness and of engagement with MOOC contents than non-completers. In addition,a logistic regression analysis indicated that the variables with greatest predictive value to discriminate between completers and non-completers were goal-setting, task interest and the academic discipline of studies of MOOC participants. The percentage of cases correctly predicted by the model was over 84%. The Spanish version of the instrument replicated the original factor structure of the SRL questionnaire and showed high internal consistency (α = 0.948).

中文翻译：

---

学习的自我调节和 MOOC 保留

摘要 背景 由于 MOOC 的高流失率，本研究旨在评估课程完成者和非完成者在自我调节学习策略和其他与 MOOC 保留相关的变量（感知有效性、MOOC 互动、动机和社会人口特征）方面的差异。这项工作还旨在翻译成西班牙语并验证一种用于评估 MOOC 中自我调节学习 (SRL) 行为的工具。材料和方法：582 名参与者回答了翻译的 SRL 问卷和其他与 MOOC 保留相关的问题。MOOC 完成者和非完成者之间的比较是在一个由 176 名本科生组成的亚组中进行的。结果：与未完成的学生相比，完成学生更有能力自我调节他们的学习，并且表现出更高的感知效率和参与 MOOC 内容的水平。此外，逻辑回归分析表明，对区分完成者和未完成者具有最大预测价值的变量是目标设定、任务兴趣和MOOC参与者的研究学科。模型正确预测的病例百分比超过84%。该工具的西班牙语版本复制了 SRL 问卷的原始因子结构，并显示出高度的内部一致性（α = 0.948）。逻辑回归分析表明，对区分完成者和未完成者具有最大预测价值的变量是目标设定、任务兴趣和 MOOC 参与者的研究学科。模型正确预测的病例百分比超过84%。该工具的西班牙语版本复制了 SRL 问卷的原始因子结构，并显示出高度的内部一致性（α = 0.948）。逻辑回归分析表明，对区分完成者和未完成者具有最大预测价值的变量是目标设定、任务兴趣和 MOOC 参与者的研究学科。模型正确预测的病例百分比超过84%。该工具的西班牙语版本复制了 SRL 问卷的原始因子结构，并显示出高度的内部一致性（α = 0.948）。