Retrieval Practice for Higher Order Thinking: Making Inflexible Knowledge Flexible

In my last blog, I outlined some of the research about retrieval practice and the ideal conditions to maximise the impact of its use. In this post, I aim to address one of the main scepticisms some people have around the use of retrieval practice, and map out a model in which the application of effective retrieval practice leads to higher-order thinking.

Sceptics would argue that simply recalling declarative knowledge, does not increase understanding, and that without having this understanding there can be no application to unfamiliar related contexts, similar to which pupils experience in their final exams. Simply asking a pupil to recall the number of deaths caused by Typhoon Haiyan will not make them an expert on tropical storms. However, effective practice, in which the level of challenge involved in our retrieval practice becomes increased, will result in this rote knowledge being transformed into flexible knowledge. I aim to outline a path in which retrieval practice moves from fact-based retrieval to more conceptual, complex and higher-order thinking.

In 2019 Pooja K Agarwal wrote a piece for the Journal of Educational Psychology, titled ‘Retrieval Practice & Bloom’s Taxonomy: Do Students Need Fact Knowledge Before Higher-Order Learning? (Link) In Agarwal’s study middle school and college students engaged in different forms of retrieval practice to determine which had the most significant impact on higher-order thinking. A sample of pupils was given no retrieval practice, only reading, a separate group undertook retrieval practice using only fact-based questions and the final group took part in retrieval practice using a mixture of higher-order and mixed quizzes. What Agarwal found was that the group undertaking retrieval practice using only fact-based questioning consistently increased delayed test performance, compared with the reading only group. The most significant finding, however, was that the higher-order and mixed quizzes group improved higher-order test performance, but the fact-based questions group did not. Agarwal concluded that “contrary to popular intuition about higher-order learning and Bloom’s taxonomy, building a foundation of knowledge via fact-based retrieval practice may be less potent than engaging in higher-order retrieval practice.”

Reading Agarwal’s study got me thinking how will the use of higher-order retrieval practice work, whilst maintaining some of the core features which are thought to make retrieval practice so effective in the first place e.g., low stakes, self-regulatory etc. Well, I may have an answer; initially, retrieval is closed and once fluency, accuracy and retention are achieved, the retrieval tasks become broader e.g., paragraph and essay writing, eventually mirroring the upper strata of Blooms Taxonomy. However, the purpose of all of my posts is to get you, the reader to consider my ideas and suggestions and reflect upon them in the context of your setting.

On first glance that might sound easier than you think, but I would argue that for what I am about to suggest to be achieved there needs to be a lot of thought and reflection into your curriculum mapping. I say this because for pupils to reach the upper strata of Blooms they need to see and articulate the links between concepts with a certain degree of sophistication. An open-ended question like ‘The Peruvian government has decided to allow the development of new roads in the Amazon, do you think this was the right decision’ requires pupils to select complex information from more than one topic area and then thread it together into a coherent argument. You might also wish to consider the argument of breadth versus depth of your curriculum. Dedicating sufficient time to higher-order retrieval practice might consume time, which would increase the depth of pupil knowledge but possibly at the expense of the breath of content to be taught. This is an argument to consider for your areas, as some will have a greater amount of curriculum time than others.

So how can we develop pupils with the cognitive dexterity to rapidly recall knowledge, identify and deploy the relevant conceptual information, all whilst using tier 2 and tier 3 language relevant to the task at hand? I will present my suggestions in a 3 step ascending hierarchy, in which the narrow and closed use of retrieval practice will become broader and higher-order.

Step One: Low Stakes Quizzing and Closed Retrieval Practice

In their 2015 book Learning as a Generative Activity (Link), Fiorella and Mayer state that for retrieval practice to be effective pupils need to receive corrective feedback and quizzes need to be taken repeatedly through spaced practice. As stated in my previous blog, feedback should be instant and self-assessed and does not need to be in the form of long and laborious written feedback. But when conducting a low stakes quiz or a similar closed, retrieval practice activity, there is the opportunity to deepen knowledge through effective questioning. As to not run the risk of this post becoming too long I would advise watching Tom Sherrington’s Rosenshine masterclass 3 on questioning for developing your questioning skills (Link). When discussing pupil responses to a low stakes quiz follow up questions can be used to move beyond simple knowledge recall to application, deepening their understanding. For example, in a quiz, when asking ‘what is erosion’ follow up questions, which require a deeper understanding might include ‘what factors affect the rate of erosion?’, ‘how is erosion different to weathering?’, ‘demonstrate your understanding of how erosion affects the shape of our coastline’. All of these questions require a pupil to have an understanding, rather than simply recalling basic facts.

Repeating quizzes through spaced practice is shown to reduce the effects of the forgetting curve. Daisy Christodoulou’s ResearchED Home (2020) presentation (Link) discusses the theory behind the forgetting curve, spaced practice and the optimum timings for spaced practice to be truly effective. What I would like to suggest here is that in the initial stages of retrieval practice it would be wise to maintain an approach of simple and factual recall with the use of corrective feedback and follow up questions. The reason for this is that it is more likely to breed success, which in turn will result in high rates of motivation. It is not motivation that results in success, but rather success that leads to motivation (Kirschner 2020). If pupils are required to instantly engage in higher-order retrieval practice, there is a high probability they will make errors and omissions that no amount of feedback will repair and inevitably will cause pupils to become demotivated. The corrective feedback is essential because without it pupils might end up cementing misconceptions in their long term memory. Imagine a golfer who practices every single day, but never has a coach to correct the errors in their technique. The longer these errors are left unchecked the more difficult they will become to undo.

I spoke about self-regulatory testing in my last post. The use of knowledge organiser and flashcards can create an environment in which pupils quiz themselves and their peers. The use of flashcards, which is also explained superbly in Daisy Christodoulou’s presentation, can be used to promote retrieval practice as home learning tasks. To begin broadening and loosening the closed nature of initial retrieval practice, consider how the questions in repeated quizzes can be overlapped. What I mean by this is that questions are rephrased. For example, in an early quiz, you might ask, ‘what is GNI per capita?’, in a subsequent quiz the question might be reversed to say ‘what is the term given to the average amount of money earned by a population?’. By overlapping questions in this manner we can begin to move away from inflexible to flexible knowledge.

Step Two: Transitioning from Closed Retrieval Practice to Higher Order Retrieval Practice

When thinking about when to move to this next step, it will be important to have reflected on this question; are the concepts that make up your curriculum been firmed up in isolation? What I mean by this is, can your pupils demonstrate accuracy, fluency and retention? This is important before we can think about asking them to integrate concepts and skills into something more complex.

The study of Agarwal revealed that disconnected, fact-checking retrieval practice will not elicit strong performance in higher-order thinking activities but that it is mixed, higher-order retrieval practice that is required to push pupils to reach the upper strata of Blooms. Mixed quizzing should contain two aspects; number one a combination of fact-based questioning and questions from the upper strata of blooms e.g., analyse, number two, questions which provide synoptic linking between topics and concepts.

Mixing fact-based questions with questions, which require further processing will require some scaffolding, especially with low ability pupils. It will be important to consider how to make each question progressively more challenging, with the higher-order thinking task supported by the less challenging, more fact-based recall questions that come before it. An example of this might look like:

  1. What is the average temperature and precipitation of the Arctic?
  2. Why might the climate of the Arctic be considered ‘hostile’? 
  3. At which latitude can the Arctic ecosystem be found? 
  4. What is the link between the latitude and climate of the Arctic ecosystem?
  5. Name one plant and one animal that live in the Arctic. For each briefly describe how they adapt to the conditions 
  6. ‘Plants and animals adapt to survive in a hostile environment.’ Explain this statement.

In the series of questions above, they progressively become more challenging but are designed in such a way that if pupils can answer the first five questions they will have success in the more challenging question six. I would describe this as long term scaffolding, which eventually will come down in step three, and it must come down, for pupils to not become too reliant on this kind of support and to promote self-regulation within learning.

I also like the idea of multiple-choice questioning in this stage and endorse their inclusion in our repertoire of questions. There are some great examples of how MCQs can be designed in a way that requires pupils to think and demonstrate their understanding. Blake Harvard’s blog (Link) has some great strategies to make MCQs incredibly effective and move beyond merely guessing the correct response from the options available.

The next aspect of mixed quizzing retrieval practice is to consider the synoptic links and design quizzes which promote rich associations and therefore build schema. Pupils need to be able to see the connections but often cannot achieve this on their own. To achieve this will require you to carefully think about how concepts are linked and then design quizzes to reveal these links. There also needs to be a focus on powerful explanations, linking where possible to prior knowledge. For example, I am currently working on a new unit of work, about economic development in India. The first lesson will explore India’s location and this provides the opportunity to forge multiple, authentic, schematic links through retrieval practice. I could explore any of the following and much more:

  1. Solar radiation and global atmospheric circulation 
  2. How solar radiation and global atmospheric circulation affect ecosystem distribution 
  3. Link this to the climatic conditions of India and as a result farming 
  4. How farming and the climate is linked to resources management 

Each of these is a concept or idea taught in isolation, contained within an individual topic, but more importantly, they are all linked. By taking this approach to mixed quizzing pupils begin to chunk information together, making the individual links explicit, organised and encoded in a more robust way into our long term memory. If we fail to explain and teach how concepts and theories fit together, most pupils will never make these links or understand how they fit together in a much larger knowledge system.

Step Three: Higher Order Retrieval Practice

The final step is to remove the scaffolding. This should happen in line with what you are seeing in your classroom. One thing I have refrained from discussing in this post is the time between transitioning from closed to more open retrieval practice and this is simply because there are some variables, which will be exclusive to each individual school and department. As practitioners, it is our responsibility to respond to our individual groups and pupils. If you observe a high degree of fluency, retention and accuracy in fact-based, closed retrieval practice it is time to move into step two. Also if you are worried because you have a class that is very mixed, and a number of pupils are being asked practice or to retrieve knowledge they have already had success in doing, therefore wasting their time, the idea of overlapping, which I explained earlier suggests that extra practice beyond the point of mastery, further aids long term memory. You could also consider ways; in which you could move individuals along the continuum at different rates. 

The final step should contain retrieval tasks, which reflects the level of challenge associated with higher-order thinking e.g., evaluative and problem-solving tasks. It should demonstrate the move from inflexible to flexible knowledge. This can only be achieved through well planned and carefully sequenced retrieval practice, which is why careful consideration of your curriculum is required. If you were to skip straight to higher-order questioning/retrieval inevitably success would be minimal, demotivating pupils and omissions, misconceptions and errors would become difficult to undo. Pupils would fail to identify conceptual links between topics and only ever be able to recall simple, factual information. 

When engaging in higher-order retrieval tasks I would refrain from merely selecting challenging questions from previous exam papers. Mark Enser in his book, Making Every Geography Lesson Count, uses the analogy of a runner, to show that someone practicing running, might improve their ability to run long distances. But that the person running will find it difficult to use what they have learnt and apply it to a different form of fitness, such as weight lifting. When considering how to best practice higher-order thinking skills, if we were to use past exam questions there is a danger that we are preparing our pupils to answer those questions – questions, which are unlikely to come up again and it will not deepen their real-life understanding of the subject you teach.

In your respected departments, I would imagine it would be easy to come up with a range of higher-order questions and activities, that aren’t taken from a previous exam. For example, when thinking about my new unit of work on India I could ask any of the following, which will require the pupil to engage in higher-order thinking and have not appeared in any previous exam paper:

  1. India is set to become the next global superpower. To what extent do you agree with this statement?
  2. Evaluate the role of outsiders, from the British Empire to TNC’s and Political allies in India’s development. 

These questions require a deep understanding of the topic at hand and the conceptual links to other topics. They require pupils to have developed flexible knowledge and will prepare them for further studying my subject at A level and University. When first engaging in this kind of higher-order retrieval practice it would be important to scaffold by using mixed and extended quizzing to help pupils identify the declarative knowledge that will underpin their response.

Final Thoughts

The advantage of the suggestions I have proposed is it will provide the long term scaffolding required to help pupils access the upper strata of blooms, with a high degree of accuracy, fluency, all whilst retaining the declarative knowledge required. Whilst I fully support Agarwal’s conclusion, that higher-order/mixed retrieval practice is more impactful than closed retrieval practice alone, I would argue that it is important initially to build the foundations of effective fact-based retrieval practice. Without the ability to recall declarative knowledge on-demand and articulate conceptual links pupils would be limited in their success when being asked to apply, analyse, evaluate etc

I aimed to set out a path, which would enable an effective transition from closed, to more broad and open retrieval practice, whilst debunking the scepticism, that retrieval practice is simply remembering facts. When carefully thought out and effectively integrated within your curriculum, retrieval practice can become so much more than a tool for simply remembering things. I believe I have been successful in achieving what I intended, but always welcome discussions, which force me to reflect on my findings and thoughts. Therefore, please get in touch if you wish to discuss the contents of this article further.

I would like to thank Blake Harvard. Blake who is a Teacher of Psychology, a contributor to the ResearchEd community and the author of the fantastic blog The Effortful Educator, kindly took the time to review and offer feedback on what you have just read. For this, I am incredibly appreciative and benefitted immensely from the experience.


Agarwal P. K., (2019) Retrieval Practice & Bloom’s Taxonomy: Do Students Need Fact Knowledge Before Higher Order Learning? Journal of Educational Psychology (Online), 111(2), pp. 189-209. Available from: Research Gate (Accessed 19th May 2020).

Christodoulou, D.,(2020) Research ED Home 2020: How to remember anything, forever. (YouTube Video) Available at:

Enser, M., (2019) Making Every Geography Lesson Count. Carmarthen, Wales: Crown House Publishing LTD.

Harvard, B. (2020) The Effortful Educator. (Blog) Available at:

Kirschner, P. (2020) Ten Tips for Emergency Remote Teaching. (YouTube Video) Available at:

Sherington, T. (2020) Rosenshine Masterclass III Questioning. (YouTube Video) Available at:

2 thoughts on “Retrieval Practice for Higher Order Thinking: Making Inflexible Knowledge Flexible

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Google photo

You are commenting using your Google account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s

Create your website with
Get started
%d bloggers like this: