Improving literacy and memory through retrieval quizzing, in schools situated in socially deprived areas 

            “Imagine, holding a hollow glass ball with walls so thin. Even with your lightest touch it bursts, shattering into a thousand pieces.” 

My mentor and head of department, describing the resilience of our students from a Cheshire town with high levels of social depravity and a growing literacy gap, during our weekly CPD.

            After moving to the town from Manchester, he described our school to me as “having the problems of an inner city without the trainers”, depicting the lack of ambition in some of our students as a result of their environment. The low levels of literacy and low resilience leading to behavioural challenges in and out of the classroom, often presenting to most as a low motivation to learn. 

Following the publication from Oxford University Press (2018), we identified literacy in science as the greatest barrier in transferring classroom ‘learning’ into success in summative assessment. This is especially true for low income pupils, who often disengage from learning due to lower reading ages and limited vocabularies. We also knew that student motivation often derives from initial success, as opposed to any sternly spoken reminders about the importance of a topic. “The examiner loves to ask this; I need everyone focused” is a particular phrase I am guilty of. Instead, we responded via embedding the philosophy of ‘resilience by design’ into our curriculum.

Instilling this ideology required us to redesign the opening 20 minutes of our lessons; we embedded “Key Term Retrieval Quizzing” into our curriculum. The start of each lesson displays the key terms for the topic arranged in a matrix on a PowerPoint slide. The teacher then describes a key term using relevant exam language, for example “I am the organelle which contains the genetic information”. The student studies a list of key terms, seeking the answer and (hopefully) displays the key term “nucleus” on their whiteboard. The entire process being repeated 15-20 times with high frequency praise, a fast pace and a growth mind-set around any misconceptions that arise.  

Lesson 2 would begin in a similar way, this time expanding the list of possible key terms. As student resilience increases, we encourage students to record their answers in their books, thus relinquishing the safety net of the mini whiteboard and allowing the possibility of failure into a medium that is not so easily rubbed away.  The teacher feeds back the first three answers, before targeting pupils to share responses and justify their reasoning.

Finally, in lesson 3 students would be presented with nothing more than a selection of key words, using the knowledge gained they freely recall appropriate definitions or explanations with minimal assistance or cues. However, I would note that low ability groups may require additional iterations of the key term retrieval before progressing to this stage. 

Free recall demands students to set aside all resources and recall as much information from memory as possible; for example, providing only one cue of “Osmosis” for students to share all relevant knowledge relating to the term. Feedback should be gathered as a class, collating information from their peers, with the expectation that all students contribute during this phase.

Sumeracki & Weinstein (2018) suggest that this form of free recall may improve learning more than cued recall, however, should students fail to recall a lot of the information, the impact on learning is reduced. Providing actionable feedback may reduce the negative impact on learning, though we should focus on improving the retrieval success first, by providing additional context to the information we desire students to recall (Smith, Blunt, Whiffen & Karpicke, 2016), perhaps through an additional review of the concept.

 It is pertinent to consider the average successful quiz score before progressing to free recall. Doing so too early would risk disaffecting low resilience pupils – too late and students may voice boredom from a lack of challenge. Rosenshine (2012) suggests 80% success or above is the appropriate time to progress as students have been sufficiently challenged and are in fact learning the material.

Our students quote retrieval quizzing to be their favourite part of their lesson, as they are initially successful, enjoy the challenge of increasingly complex key terms and are no longer discouraged by new language when it is presented.  The process of removing the safety net of the key term matrix from quizzing, however, was another challenge entirely. 

The evidence supporting the use of quizzing in classrooms as an effective consolidation strategy is strong. When re-exposing students to concepts and knowledge during retrieval quizzing, we, in turn, space learning over time.  By providing time for students to forget presented information we can distribute practice of an idea or skill (Pashler et al, 2007), no matter how frustrating it can be to observe your near-perfect explanation of an idea be completely lost a week later!

 The storage strength of an idea reveals how embedded it is in our long-term memory. For example, the postcode of your childhood address would have a high storage strength, as you have recited this many times. The retrieval strength of an idea, on the other hand, is how accessible that memory is (Bjork & Bjork, 1992), perhaps what you ate for lunch today.

The limitations of our practice stemmed from the reduced efficacy of quizzing from multiple choice questions. Although the options are much greater than standard multiple choice, the process of completing a mental ‘cloze activity’ may support students in narrowing the solution down, i.e. to a verb for a process or a noun for scientific apparatus. This process did not always lead to effortful retrieval of information; therefore, our students were not always making the necessary gains in storage strength. 

As a faculty, we became increasingly dependent upon the key-term grids to ensure initial success, perhaps inflating our own egos with the number of 20/20 quizzes achieved, or to simply ensure students were sufficiently motivated and to minimise challenging behaviour from a collapse of resilience. In short, we needed to increase the level of challenge by reducing our cues during questioning, without curbing the motivation of our students. 

In response, we explained to our students the importance of effortful retrieval from memory. After expanding the key-term grids, we then began to remove them. This was initially met with reluctance from the students, but over time, the students relished the challenge. 

One group of triple science students have not seen a prompt grid in months – and take great pride that fact. Another Year 8 group were found boasting to their history teacher of their ability to recall quark structures of hadrons and their understanding of the Pauli Exclusion Principle, as a result of our literacy quizzing. Year 7 students were also heard bragging that their class could use the word “pluripotency” correctly when describing stem cells. A culture of challenge and competition between science groups has ensued ever since. 

Surprisingly, there was little variance between “cued” quizzes and reduced cue quizzes. In a control group of year 9 triple science, the variance was as low as 9%, thereby having little statistical significance. We were able to vary the challenge through increasing the spacing of our questions over time and re-instating the cues when beginning a new topic, before repeating this “phased-cue reduction” process again. 

The results revealed that a cohort of students may transition from recognising scientific terminology from a list of responses, to retrieving information from fewer cues. Hence, low-resilience students are encouraged to depend on their long-term memory. Phased cue reduction may effectively contribute to an increased reliance on memory, without damaging motivation to effortfully retrieve information.

The future for us lies in converting our student’s improved scientific literacy into extended responses and closing our numeracy gap.  We are encouraging our students to convert answers from key term quizzing into their extended practice questions around fundamental ideas in science and gradually embedding standard form conversions, percentage calculations and unit conversions into our key term quizzes.  

Every minute is precious in lessons, and balancing time spent delivering content with that spent retrieving information is a problem we are striving to resolve. There may be a golden ratio between retrieval and delivery to be found but, like our students, we still have much to learn. 

Lewis Stewart is an Expert Teacher at See the Mountain, operating in Cheshire, Staffordshire and Greater Manchester

References

Bjork, R. A., & Bjork, E. L. (1992). A new theory of disuse and an old theory of stimulus fluctuation. From learning processes to cognitive processes, Essays in honor of William K. Estes, 2: 35-67

Didau, D., & Rose, N. (2016). What Every Teacher Needs to Know About Psychology. JOHN CATT EDUCATIONAL LTD.

Dunlosky, J. (2013). Strengthening the pupil toolbox: Study strategies to boost learning. American Educator, 37(3): 12-21

Pashler, H., Bain, P. M., Bottge, B. A., Graesser, A., Koedinger, K., McDaniel, M., & Metcalfe, J. (2007). Organizing Instruction and Study to Improve Pupil Learning. IES Practice Guide. NCER 2007-2004. National Centre for Education Research

Sumeracki, M. & Weinstein, Y. (2018) Optimising learning using retrieval practice. Impact. 2: 13-16

Are you sitting comfortably? Then we’ll begin… Story time may be among the least likely things you’d expect to find in a science class, but when Lewis Stewart and his colleagues realised students’ working memories were being overloaded with complicated ideas, they decided to turn to the power of spinning a yarn to help weave meaningful learning… 

What was the problem you were finding in science lessons?
There are often instances in chemistry – as well as in other subjects – where our students are required to commit multiple stages in a process to memory. Often such processes require our students to hold multiple “interacting elements” in their working memory at any given time.

When we ask students to explain how two elements can chemically combine in an ionic bond, for example, they must consider the electron structure for each element, how many electrons may be transferred, what effect this has on the atom’s charge and the properties of the new compound formed. You can can see how this might overwhelm students, and quite often, such tasks did.

We also identified a further problem that a lot of our explanations and modelling in science had to be delivered at distance from the students and often in rooms unsuitable for many science demonstrations. Therefore, we had to consider a way of improving the quality of our explanations and modelling of abstract concepts, without exceeding our students’ working memory capacity.   

What sort of impact was this having on students?
Whenever task challenge is too high, or when our students lack the required prior knowledge to manage incoming information, we risk damaging their perceptions of their own competence or providing negative learning experiences that may leave them feeling less motivated. Our school is situated in a disadvantaged area, and some of our students have low levels of resilience. 

We hope to leave our students feeling successful after every lesson, but we cannot shy away from delivering the more rigorous and challenging aspects of our science curriculum. Delivering this ‘powerful knowledge’ – which takes our students beyond their lived experience – is our duty as their teachers, and their greatest chance in increasing their opportunities.

What made you think about storytelling as a possible remedy?
Reflecting on my experiences teaching the history of the atom, my students could visualise the University of Manchester’s buildings as “the birthplace of nuclear physics” after I explained how physicist Ernest Rutherford’s desk still remains irradiated to this day. This nestled the core knowledge of his observations and conclusions within a memorable story. 

I feared the story was extraneous, or that I was perhaps wasting precious time, but the durability of my students’ knowledge, evident in later retrieval quizzing, suggested the story helped enforce meaning and made Rutherford’s alpha-scattering experiment even more memorable. 

After reading literature from Daniel Willingham, I thought the durability of my students’ knowledge may be a result of the “privileged status of story in memory” (Willingham, 2004). Literature suggests anchoring new ideas in stories may increase motivation and provide a strong base for students to return to if they later forget (Pashler et al., 2007). 

Research from the past 30 years suggests stories are specially placed in memories, referred to as “psychologically privileged”. Such stories contain the 4 Cs:

1.       Causality: related events in a story

2.       Conflict: obstacles preventing a goal from being met

3.       Complications: new problems which must be solved

4.       Characters: observing characters in action

Stories that require readers to infer meaning through making what are known as “medium-level inferences” are deemed more interesting (Willingham, 2004). This supported my decision to evaluate how storytelling may develop conceptual understanding in chemistry.

When implementing stories in my classroom, I had to carefully consider the degree of causality and the level of inferences my students must make. Evidence suggests stories with medium-level inferences are also most accurately recalled (Keenan et al., 1984). 

Therefore, developments in the story must be sequenced to ensure students can draw their own connections between related events. Willingham (2004) proposes stories provide a non-threatening and interesting introduction to new concepts  (Willingham, 2004).

So how does this work in practice?
If the intended knowledge is difficult to model and abstract with lots of interacting elements, a story may be useful. If the knowledge contains a series of events or stages in a process, influencing one another in turn, then a story with high causality may support students in recalling each stage. 

We made a magnesium atom into a character, who was tired of carrying around two extra electrons, this represented the conflict and created a goal for the story. Another character, a fluorine atom, offered a solution to the goal, by offering to take an electron from magnesium. A complication was introduced, the fluorine atom could only take one of magnesium’s electrons. 

However, the solution of fluorine’s twin brother arriving and offering to take an electron provided a memorable solution to the problem. The causal events in the story made each step in the process explicit and memorable.

If the knowledge requires students to evaluate, compare or contrast ideas, the story should be adapted. When students were tasked to compare processes of purifying water, they recalled more of the stages in each process and their respective advantages and disadvantages when we had a character set in a familiar location, trailing each method and facing challenges along the way.

It is also important to note students must be invested in the story’s characters, enveloping them in the inquiry process. To achieve this, the teacher may consider making the conflict and complications “affectively charged” – capturing topics relevant to students, such as family, friends, and identity formation (Isabelle, 2007, p.23).

What sort of response have you seen from students so far?
After introducing stories into our science curriculum over multiple groups, we observed an increase in student performance on extended response questions relating to each story and increased performance on independent practice questions. Students retrieved stories with an accuracy of 87 per cent after a delay.

Six months on from story one, students in Year 10 can still fully explain ionic bonds between elements. In an end-of-unit assessment, every student successfully explained the ionic bond between two atoms. Although our stories were only recently introduced and more rigorous studies should be conducted before making any firm claims of their success, the early indications suggest storytelling in science is improving student retrieval and performance in extended answer questions.

Lewis Stewart is an Expert Teacher at See the Mountain, operating in Cheshire, Staffordshire and Greater Manchester