“If you want to know the optimal distribution of your study time, you need to decide how long you wish to remember something”
Cepeda et al. (2008, p. 1101)
Jade Pickering, University of Southampton
Julie Hadwin, Liverpool Hope University
Phil Higham, University of Southampton
Kou Murayama, University of Tübingen
Rosalind Potts, University College London
In years past, educators have lamented the fact that research in cognitive psychology was providing no real answers to important real-world questions such as “What is the most efficient way for students to learn information to create durable memories?” (e.g., Dempster, 1988). However, there is evidence that this situation is changing. For example, investigations of highly effective learning techniques such as spaced practice (spreading learning over time; e.g., Emeny et al., 2021) and retrieval practice (quizzing of learned information; e.g., Yang et al., 2021) in real classrooms are increasing exponentially.
Successive relearning combines spaced practice and retrieval practice into one methodology. Evidence suggests that it strongly promotes durable memory (Rawson et al., 2018). More recently, Higham et al. (in press) found that it also reduced the learner’s worry and increased their self-efficacy.
Further research on successive relearning is, however, needed to apply it successfully to real classrooms. Specifically, Cepeda et al. (2008) found that there is an optimal spacing gap for learning that depends on how long information needs to be retained. The longer individuals are asked to remember information, the longer the spacing gap should be. However, if the spacing gap between learning sessions is too long, between-session forgetting makes practice ineffective. This finding is important because it suggests that different schedules are optimal if students are learning for an exam in a few weeks versus month or years.
The aim of this project is to identify relearning schedules that optimise student learning outcomes across different learning timelines. We extend existing research by considering the relationship between the number and timing of practice sessions for relearning material, in the context of different retention intervals associated with its recall. Having multiple practice sessions is more like real-world education and should affect decisions about optimality, making it critical to investigate. We anticipate that multiple practice sessions should reduce forgetting, which, in turn, will lengthen the optimal spacing gap. We will also consider individuals thoughts and feelings about their learning journey, to explore e.g., their predictions of their future recall accuracy, and feelings of anxiety, attention, and self-efficacy.
Participants recruited via Prolific will learn vocabulary over multiple practice sessions using English – Swahili word pairs (e.g., bustani - garden ). Some word pairs will consist of retrieval practice where participants are asked to recall the English translation to a Swahili word ( bustani - ? ). To control for time-on-task, we will ask participants to restudy other word pairs (i.e., bustani – garden ). To achieve mastery, we will repeat word pairs within a single session, as well as across sessions, with corrective feedback following each retrieval attempt. We will gather participant judgements of learning, as well as ratings of anxiety, attention, and self-efficacy at the end of each practice session.
Thirty-six independent groups of participants will undergo initial learning, and will experience one or more practice sessions, across different learning timelines and will take a final test. We will manipulate: (a) the number of practice sessions following initial learning (1, 3, or 5), (b) the retention interval (time between the final practice session and the test: 7, 35, or 70 days), and (c) the spacing gap (time between practice sessions: four levels). We will include three levels of spacing intervals across the levels of the other factors (0 [massed], 1, and 50 days). A fourth level of spacing interval was obtained from Cepeda et al. (2008) and varies with the retention interval: 3, 8, and 12 days for 7-, 35-, and 70-day retention intervals, respectively. Participants will therefore be enrolled for 8-321 days.
Using a 3 x 3 x 4 ANOVA, we will examine the effects of the number of sessions (1, 3, or 5), the length of the retention interval (7, 35, or 70 days), and the length of the spacing interval (massed , short [1 day], medium [3, 8, 12 days], long [50 days]) on accuracy at final test. We will follow this up with post-hoc simple interaction and main effect analyses. We expect spacing interval to interact with the retention interval, but that this may be moderated by the number of sessions.
Additional exploratory analysis will look at the effects of restudying compared to retrieval practice, and learners’ thoughts and feelings about the learning process.
An a priori power analysis indicated that the sample size needed to detect a small-to-medium size effect (Cohen’s f = .20; power = .80; alpha = .05) for a 3-way interaction between spacing interval (4 levels), retention interval (3 levels) and practice sessions (3 levels) is 445 total participants.
Each session (encoding, retrieval practice, and final test) will take approximately 30 minutes, on average. The number of sessions and the likelihood of attrition (and subsequent need to replace participants) will vary depending on participant group.
We will pay participants £5 per hour. Groups 1-12 will participate for approximately 90 minutes total (£7.50), Groups 13-24 will participate for approximately 150 minutes total (£12.50), and Groups 25-36 will participate for approximately 210 minutes total (£17.50). We will over-recruit (40 total participants) in the 6 groups with the longest schedules, and recruit on-target for the groups where we can (timewise) afford to enrol a new cohort of participants after data collection to address attrition. Actual attrition has been estimated to need 20 participants total for the remaining groups. Thus, the total cost (including a 33% service fee and VAT [20% of service fee]) is estimated as £15,120.
We are applying for the full £10k grant and will aim to cover the remaining costs from an existing grant.
The study is pre-registered on the OSF. We will make study materials and (anonymous) data publicly available on the OSF and analysis code available on GitHub upon study completion. The publication will be available as a pre-print and published open access.
This study has been approved by the Ethics Committee at the University of Southampton (Reference: 57458).
Cepeda, N. J., Vul, E., Rohrer, D., Wixted, J. T., & Pashler, H. (2008). Spacing effects in learning: A temporal ridgeline of optimal retention. Psychological Science, 19 (11), 1095–1102. SAGE Journals: Your gateway to world-class journal research
Dempster, F. N. (1988). The spacing effect: A case study in the failure to apply the results of psychological research. American Psychologist , 43 (8), 627–634. APA PsycNet
Emeny, W. G., Hartwig, M. K., & Rohrer, D. (2021). Spaced mathematics practice improves test scores and reduces overconfidence. Applied Cognitive Psychology, acp.3814. https://doi.org/10.1002/acp.3814
Higham, P.A., Zengel, B., Bartlett, L., & Hadwin, J.A. (in press). The benefits of successive relearning on multiple learning outcomes. Journal of Educational Psychology . https://doi.org/10.1037/edu0000693
Rawson, K. A., Vaughn, K. E., Walsh, M., & Dunlosky, J. (2018). Investigating and explaining the effects of successive relearning on long-term retention. Journal of Experimental Psychology: Applied , 24 (1), 57. APA PsycNet
Yang, C., Luo, L., Vadillo, M. A., Yu, R., & Shanks, D. R. (2021). Testing (quizzing) boosts classroom learning: A systematic and meta-analytic review. Psychological Bulletin , 147 (4), 399–435. APA PsycNet