Helen Skelton, Head of Science, Beaumont School, UK
Knowledge is required to think and therefore to learn. Learning has been defined as ‘relatively permanent changes in comprehension, understanding, and skills of the types that will support long-term retention and transfer’ (Soderstrom and Bjork, 2015, p. 176). The long-term retention of learning is a particular challenge in mathematical aspects of Key Stage 5 chemistry. Students are frequently able to answer problems soon after instruction when they all involve the same, recently learned, concept. However, they find it more challenging when problems are presented after a time delay or in a mixture, requiring recall and selection of the appropriate approach.
The evidence
Dunlosky et al. (2013) assert that the study strategies commonly used by students are not the most effective. My students are resistant to exploring alternative study strategies, probably due to the illusion of knowing (Bjork et al., 2013) that popular strategies, such as rereading and highlighting, can elicit. However, the extensively investigated alternatives of practice testing (or retrieval practice), distributed practice and interleaved practice are strategies that could be applied to improve consolidation (Bjork et al., 2013; Perry et al., 2021), particularly in Key Stage 5 chemistry (Hartman et al., 2022).
Exploration of research identified several factors as important in implementing spaced practice testing, combined with interleavingAn approach to learning where, rather than focusing on one piece of content at a time (known as blocking) then moving on to the next, students alternate between related concepts (summarised below). A preliminary intervention was designed (see section, ‘The Intervention’), the implementation of which enabled refinement of the identified key ingredients, summarised below, with italics indicating alterations following the preliminary intervention:
- Identification of content (Dunlosky and Rawson, 2015): Core knowledge is clearly identified and shared with students
- Low-stakes (Adesope et al., 2017): A tool for consolidation and not a test
- Knowledge retrieval (Perry et al., 2021): Promoted through classroom administration
- Mixed, generative question format with content similar to final tests (Dunlosky and Rawson, 2015; Adesope et al., 2017; Yang et al., 2021): A mix of knowledge recall and mathematical questions
- Time delay (Wiseheart et al., 2019; Latimier et al., 2021): Retrieval gap in the order of weeks when feasible within classroom constraints
- Content is different but related to current learning: Exploits some benefits of interleaving (Firth et al., 2021)
- Variation of context (Bjork and Bjork, 2011) adds a desirable difficulty
- Corrective feedback that explains errors (Dunlosky and Rawson, 2015; Yang et al., 2021): Students understand their mistakes, which supports accurate future recall; enables students to identify areas for additional study and practice.
The intervention
Having identified evidence-informed principles relevant to the problem of consolidation, I implemented the following intervention (italics added following the preliminary trial of intervention):
- Identify and share core knowledge (including relevant prior knowledge from Key Stage 4), problem types and solution procedures
- Begin lessons with knowledge quiz completed independently, from memory – quiz contains recall (at start of topic) and mathematical problems (added later). (See Figure 1)
- Review answers – students self-correct recall questions and check answers to mathematical questions; teacher circulates and reviews answers during quiz, allowing tailored individual and whole-class feedback
- Teacher reviews any incorrect responses during the lesson
- Teacher gives feedback on mathematical question to the class or individuals
- Reteach any areas where recall or application is less secure; identify any additional core knowledge to add to summary, reteach or provide additional practice
- Students identify areas for independent study or practice.
Figure 1: Exemplification of questions in knowledge quizzes
Results and discussion
The intervention revealed that repetition of poorly answered recall questions across multiple quizzes resulted in increased correct responses, indicative of the importance of retrieval, spacing and corrective feedback (Perry et al., 2021; Yang et al., 2021). Feedback was particularly important, as students initially recalled the gist of definitions but omitted important specific details.
Varying the context for practising recall adds a desirable difficulty (Bjork and Bjork, 2011). Evidence of this challenge emerged from examples where students were able to correctly answer a simple recall question but were unable to recall the same knowledge in the context of a problem without practice. Thus, it is important to include opportunities for students to recall knowledge in the context of unfamiliar problems.
Following five weeks of the intervention, students completed an end-of-topic test. The intervention cohort (n=12) scored significantly more highly (95 per cent confidence interval) than the five previous cohorts combined (n=70) (Figure 2). Comparable outcomes were expected, as the current cohort performed comparably to previous cohorts in a baseline assessmentWhen referring to early years education, a measurement of a child’s performance conducted within a few weeks of them starting school in Reception, with a focus on literacy and numeracy.. Although encouraging, this improvement may not be solely attributable to the intervention, as students have received greater individual teacher attention, with the class size being 56 per cent of the cohort mean.
Figure 2: Comparison of test results between students who have and haven’t undergone the intervention
In addition to improved test performance, it was hoped that the intervention would boost students’ confidence in tackling unfamiliar problems. Students’ ratings of their confidence in different problem types increased after the intervention. Although high confidence did not correspond to success in particular problem types in the end-of-topic test, increasing confidence did correlate more strongly with success in practice quizzes. This may be due to the metacognitive effects of highlighting areas where knowledge is less secure, through practice testing with formative feedback (Roediger et al., 2011), which informed more focused independent learning immediately following the practice quiz.
A third desired outcome was improving students’ resilience in persevering with solving problems from memory that they could not immediately answer. This ensures that students retrieve as much as possible to maximise the benefits to long-term memory (Perry et al., 2021). Although students displayed improvement in their ability to solve all types of mathematical problems independently, this has not affected the way in which most approach problems outside of the classroom (frequently relying on notes or peers, rather than persevering with recall), reinforcing the importance of classroom administration to secure the benefits of retrieval (Yang et al., 2021).
Conclusions
Although limited in scope and time, this intervention has highlighted some important principles for successful consolidation of learning through spaced retrieval practice, in addition to those identified from research:
- Classroom retrieval has added benefits – the teacher can enforce recall from memory and tailored feedback can be provided
- Retrieval may be more beneficial if unfamiliar contexts are included as well as straightforward knowledge recall
- Students often interpret almost correct answers to be correct – even for simple recall
- Retrieval supports students in identifying strengths and weaknesses for independent study.
The work described was completed as part of my Master’s in Expert Teaching, delivered by Ambition Institute in partnership with Plymouth Marjon University.