Effective learning hinges on recalling and applying prior knowledge. Research demonstrates the "spacing effect," where distributing learning over multiple sessions improves long-term retention compared to massed learning (single, lengthy session). This effect has been consistently replicated across various domains and age groups in laboratory settings. However, most studies control learning environments, simplifying stimuli and limiting learner agency. Real-world learning scenarios, particularly in educational settings, involve self-regulated learning and more complex stimuli. The current study addresses this gap by examining self-regulated spacing in a real-world MOOC setting. It investigates whether the spacing effect translates to complex, self-directed learning, considering the interplay of learner control, and different learning styles, specifically relating study time distribution and practice activity completion rates to learning outcomes.

The spacing effect, initially observed over a century ago, is a well-established phenomenon in cognitive psychology. Early research, using simple stimuli like word lists, demonstrated that distributed practice consistently leads to better long-term retention than massed practice, particularly when tested after a delay. While research consistently shows the advantages of spaced study, most prior studies focus on controlled laboratory settings, often manipulating spacing directly and using relatively simple stimuli. There is limited research investigating self-regulated spacing behaviors within naturalistic educational settings, where learning is more complex and learners are in control of their study strategies. Existing research shows that even when spacing is prescribed in a language class, benefits are observed compared to massed practice. However, the extent to which self-directed learning and its impact on how spacing affects learning outcomes remains an important gap in current literature. Previous studies suggest that the learner's control over their study pace and spacing can significantly influence the relative benefits of spaced practice compared to situations with less learner control.

This study analyzed data from a 12-week introductory psychology MOOC offered on Coursera. The dataset, sourced from DataShop, an open learning repository, included data from 437 students who completed the course, encompassing 11 weekly quizzes and a final exam. Key variables included pretest scores (measuring prior knowledge), learning outcomes (quiz and exam grades), and study behaviors (spacing, study time, retention interval, and activity completion rate). Spacing was defined as the number of study sessions per unit. Study time was calculated as the total duration of all sessions. The retention interval measured the time between the last study session and the unit quiz. Activity completion rate tracked the number of integrated activities completed per unit. Regression analyses were used to examine the relationships between spacing and quiz performance, controlling for pretest scores and total study time. Random effects models accounted for variations between units and students. Interactions between spacing, retention interval, student ability (proxied by final exam score), and activity completion were also explored. Data visualizations (histograms and plots) were used to illustrate the distributions of study behaviors and the relationships between variables.

The analysis revealed a significant positive correlation between self-regulated spacing and unit quiz performance, even after controlling for total study time and retention intervals. Students who spaced their study across more sessions achieved higher quiz grades. While there was also a main effect of retention interval (shorter intervals leading to higher quiz scores), there was no significant interaction between spacing and retention interval, indicating that spacing benefits were present irrespective of the time between study and test. Similar results were observed for the final exam grade. Further analysis showed that higher-ability students (as measured by final exam scores) were more likely to employ spaced study. However, the beneficial effect of spacing on quiz performance was more pronounced for lower-ability students. Similarly, students completing more course activities tended to use spaced study, but the relationship between spacing and quiz grades was stronger for students who completed fewer activities. This suggests that spacing may act as a compensating factor for students who do not engage extensively in active learning strategies. This was confirmed via a regression analysis predicting quiz performance using both spacing and number of activities completed. A significant interaction effect was found showing that when students completed fewer activities, the positive relationship between spacing and quiz performance was more substantial.

The findings indicate that self-regulated spacing is a valuable learning strategy, even in naturalistic learning contexts where learners manage their study independently. The positive relationship between spacing and learning outcomes is particularly notable given the complexity of the learning environment and the lack of direct control over spacing by the researchers. The observation that the benefits of spacing are most evident for lower-ability students and those who do not frequently engage in active learning practices suggests that spaced study might act as a compensatory mechanism. This may be because both activity completion and spacing promote retrieval, and if one strategy is already in place, the additional benefit of another strategy that utilizes the same mechanism would be less than additive. The data shows that students who don't engage in many activities can use spacing to compensate and still see high learning outcomes. These results highlight the importance of considering individual learning strategies when designing effective learning interventions.

This study demonstrates the effectiveness of spaced study in a real-world online learning environment, even when spacing is self-regulated. The benefits of spacing were particularly pronounced for lower-ability students and those who did not engage extensively in active learning. Future research could explore the underlying mechanisms mediating the relationship between spacing and learning outcomes, potentially focusing on the role of retrieval practice and individual differences in learning styles. Examining the interaction of spaced study with other effective learning techniques would be beneficial to further optimizing learning strategies.

The study relies on observational data from a specific MOOC, which may limit the generalizability of the findings. The use of final exam scores as a proxy for ability is a potential limitation, as it may be correlated with spacing behavior itself. Furthermore, the study did not control for other factors that could influence learning outcomes, such as motivation, prior educational experience, or individual differences in learning preferences. Therefore, the causal inferences from this study need to be carefully interpreted.