See discussions, stats, and author profiles for this publication at: https://www.researchgate.

net/publication/234665205

Working Memory and Learning

Article in The School science review · January 2010

CITATIONS READS

9 833

1 author:

John Crossland

20 PUBLICATIONS 49 CITATIONS

SEE PROFILE

All content following this page was uploaded by John Crossland on 18 February 2015.

The user has requested enhancement of the downloaded file.

 Working memory and learning
John Crossland

ABSTRACT Feedback from teachers during in-service courses shows that they are fascinated by
neuroscience, as they feel that it has the potential to improve their teaching practice. A previous
article of mine in School Science Review reported the first eight messages of the outcomes from
a small-scale action research project with primary and secondary science advanced skills teachers
(ASTs) in North Yorkshire. The second in this two-part account introduces a further seven messages
related to aspects of how the human brain learns that provide credence for the ASTs’ exemplary
practice. Finally, the 15 messages are grouped together into a series of concluding notes to help
teachers to improve pupils’ learning in their classrooms.

A previous article, ‘Brain biology and learning’ about the lesson introduction, collaborative group-
(Crossland, 2010), introduced what we called the working and the review/plenary;
‘big-picture model’ of the brain and the first eight l possible areas for further research.
messages from an action research project. It also
gave the background to the six years of action Working memory models
research carried out in North Yorkshire by primary Working memory is a theoretical framework from
and secondary science advanced skills teachers cognitive psychology that refers to the ability of
(ASTs). The summary from ‘Brain biology and the brain to hold and manipulate information in
learning,’ stated that there is a necessary hierarchy the mind for a short period of time (in the order of
for high-quality learning: seconds). Positron emission tomography (PET)
and functional magnetic resonance imaging
Successful learning depends first upon an
(fMRI) research has shown that working memory
emotionally positive context that enables thinking
is a widely distributed network of associated brain
processes to take place in order to produce the
areas mainly in the neo-cortex but also including
learning outcome.
the cerebellum. There is substantial evidence that
This is in stark contrast to learning as defined working memory plays an important part in
by the National Strategies for England, where learning; for example, measures of working
the emotional setting and thinking processes are memory are closely related to performance in key
not usually explicit in the lesson and therefore stage test results, that is low scores for working
teachers may consider them to be less important memories are associated with low test scores.
than the explicit learning outcomes (WILF – What Studies with ADHD pupils (Klingberg et al.,
I’m Looking For). 2005) suggest that training the working memory
Also, the human brain has a built-in capacity increases the range of cognitive abilities with a
to work collaboratively in social groups to solve resultant increase in IQ test scores.
problems that individuals may not be able to solve
on their own. Message 9. Working memory is an important
This account includes: factor in determining success in learning
in classrooms. Therefore teachers need to
l another seven messages for teachers,
devise tasks that explicitly take account of the
numbered from 9 onwards, to provide continuity
limitations and strengths of working memory. For
from ‘Brain biology and learning’ (Crossland,
example, teaching pupils to think about breaking
2010);
down tasks into smaller units and to make
l two different models for working memory;
written notes (or use the dictaphone in a mobile
l conclusions based on bringing together some
phone) of the progress towards a solution.
of the messages to provide teachers with guidance
SSR September 2010, 92(338) 105
Working memory and learning Crossland

The capability of working memory has two of the representations (memories of knowledge,
features, although there is much disagreement understanding, skills and life events) in long-term
about the definition of each. The first feature is memory. There is no limit to the activation of
processing speed and the second is storage representations in long-term memory but in adults
capacity. These both increase up until the age of only four can be the conscious focus of attention at
about 15 when pupils achieve adult levels of any one time. Oberauer (2002) extended Cowan’s
performance. The ±2.5 year variability in the model by proposing that one of the elements is also
development of working memory for school able to manipulate the information in the other three,
pupils of the same age is striking. Another is the therefore having an executive function (Figure 1).
catastrophic loss of information caused by Sometimes when pupils are in a group, their
interruptions to attention. Even if there is no working memories do not collectively engage in
catastrophic forgetting, interruptions can cause the cognitive task, they are just sitting together
both the encoding and recall of memory to be (see Figure 2a). When the task is sufficiently
distorted, leading to future faulty thinking and challenging, the pupils can work collaboratively by:
attention biases. 
l refreshing each other’s natural forgetting and
beginning to construct new thinking through a
Message 10. Working memory is prone to
process of exploratory talk (Mercer, 2000);
forgetting or corruption due to distractions. and sharing some of the representations
l using

Teach pupils to be aware of their distractive
in their working memory that are not fully formed
behaviours by emphasising listening and taking
for expert performance, that is the representations
turns before presenting their ideas to others.
still have a zone of proximal development
Teach pupils to use conscious repetition to
(Vygotsky, 1978).
renew their own working memory for speech
and visual stores. The result of the group working collaboratively is
that the pupils’ individual zones of proximal
Several models have been proposed for
working memory and some of the earlier ones
referred to short-term memory, but this term has (a)  
been superseded or included in the term working
memory. There are two accepted models that
have some important implications for learning  
in classrooms. Cowan’s model is discussed first  

because it has a simple structure. The second was

proposed by Baddeley and Hitch in 1974 and
revised by Baddeley in 2000.
Cowan’s model (2001)

Cowan regards working memory not as a separate
system but as a part of long-term memory.
Representations in working memory are a subset
(b)  



 
 




 

Figure 1 Oberauer’s extension of Cowan’s individual Figure 2 Oberauer’s extension of Cowan’s working-
pupil’s working-memory model memory model and group work

106 SSR September 2010, 92(338)

 
Crossland Working memory and learning

development appear to combine to form an concerning the basic structure and function of the
additional focus of attention, that is the group model (Meulen, 2008).
appear to grow a temporarily larger working
memory (Figure 2b) and the group has the Message 12. Pupils will only remember a
cognitive capability to solve problems that are just teacher’s instructions verbatim if they are about
too difficult for the individual working memories. two seconds in length and the last thing said to
This effect is multiplied for groups of three or four. them. Anything else has to be processed and
Our experience is that, for larger group numbers, encoded so that it can be stored temporarily
pupils in secondary school rarely have the elsewhere (e.g. long-term working memory –
group-working skills to ensure that all members of see below). Good strategies to improve pupils’
the group are taking part in a collaborative way, so understanding of the purpose of a task is for
there is no additional benefit. teachers to provide clearly thought-out spoken
instructions and, if they are longer than two
Message 11. Cognitively challenging tasks seconds, teach pupils to convert the instructions
set in the context of collaborative group work into a short list of key words, a meaningful
enable small groups (2–4) of pupils to learn diagram, a series of pictures, a flow diagram or a
more effectively how to achieve a solution than story of actions.
they could as individuals. The teacher can
achieve this without the need for intervention, Another recent proposal is that the inner voice,
for example by teaching the pupils to ask each through the process of talking to oneself, has a
other for help before asking the teacher. controlling influence over attention and behaviour
over and above that of the central executive.
Baddeley’s model (2000)
The original Baddeley and Hitch model from 1974 Message 13. For challenging tasks, a pupil
places working memory as a separate system from talking aloud is a normal response to a working
long-term memory and it has multiple components, memory at its limit of capacity. Further, the talking
for example a central executive that manipulates the seems to help with concentrating on the task,
three stores for information (Figure 3). The three releasing working memory space in order to
stores are for speech, visual and spatial information. solve tasks using skills that are only just being
The model has two slave systems for the consolidated, and finally it has a regulatory
maintenance of current information: the inner function over behaviour. Teachers and pupils need
voice (phonological loop) and the mind’s eye to recognise the importance of this type of talk as
(visuo-spatial sketch pad) (Figure 4). The inner a natural part of resolving difficulties. If the talking
voice manipulates speech and maintains the causes interruptions to others then the pupils
speech store. Only about two seconds’ worth of could be encouraged to talk ‘inside their head’.
speech can be stored and it rapidly decays (a few
seconds). Longer sections of speech need to be The model was updated (Baddeley, 2000) to
processed down to essential meaning and stored, contain an additional store, the episodic buffer
sometimes in other temporary stores – see below. (Figure 5). The buffer holds links from the
The mind’s eye generates, manipulates and current working memory to active representations
maintains the spatial (physical location) and visual in long-term memory that are available both
stores. The mind’s eye storage is more persistent consciously and pre-consciously, that is previously
than that of the inner voice. They can both convert learnt procedures and personal routines are very
each other’s information into their own form, thus quickly called upon to support current working-
releasing the other store for additional storage. memory processes. This is sometimes called
The central executive is, among other things, long-term working memory (LTWM). Recent
responsible for directing attention to relevant fMRI findings indicate that the medial temporal
information, suppressing irrelevant information lobes of the neo-cortex play a crucial role in
and inappropriate actions, and for coordinating both long-term memory encoding, and working
cognitive processes when more than one task has memory, indicating that they work together and
to be processed at once. There is strong empirical therefore support the concept of an episodic
support from psychology and neuroscience buffer. As a result of the episodic buffer linking
SSR September 2010, 92(338) 107
Working memory and learning Crossland

 




       

  



Figure 3 The structure of Baddeley’s original model

 



    


      

  



Figure 4 The function of Baddeley’s original model

to LTWM, the storage capacity and processing provides more pre-conscious representations
power of working memory is temporarily that shorten the time for decision-making, with a
increased. The effectiveness and efficiency resultant dramatic and effortless improvement in
of the links in the buffer are task specific. For performance for that specific activity. Through
example, as an individual pupil acquires expertise the process of practice makes perfect, pupils can
in understanding a series electrical circuit, be expert performers in some tasks while being
LTWM is developed to expand the capacity and novices in tasks of similar cognitive difficulty.
processing limitations of working memory. These In the classroom, pupils with high levels of
new representations are then used to understand expertise retrieve information predominantly
increasingly more complex parallel circuits. In pre-consciously for skills, knowledge and
this way, expert practice provides a superior range understanding. For example, to understand
of activity-specific representations for focusing photosynthesis in a leaf, pupils build up an
attention on the most important elements of understanding of the internal structure of a leaf
the data input to the brain. Expert practice also and this can become an active representation in
108 SSR September 2010, 92(338)
Crossland Working memory and learning

Figure 5 The structure and function of Baddeley’s (2000) working memory model (adapted)

LTWM. Next, they can consider the function of Now think about the pre-conscious representations
each part of the leaf and this knowledge becomes and routines you used that made the text
available for active representation. Once this level surprisingly easy to read. It has taken you years of
of understanding is reached they can use these practice to build up banks of representations to
active representations to answer questions about achieve this level of reading expertise.
the gas exchanges and chemical reactions taking
place. Lower levels of expertise do not have these Message 14. Less-expert pupils, with fewer
active representations available so the thinking pre-conscious representations for a task, require
process can only take place under slower more time to reconstruct their answer to a
conscious control. The current brain-based view is teacher’s question. For the typical response
that the vast majority (estimated at 98%) of the time of less than two seconds in the secondary
brain’s processing uses these pre-conscious classroom, only the experts will be able to
routines to process the immense amount of answer because they have time to retrieve
information entering the brain, and only about 2% the answer pre-consciously. Novices will
is routed to working memory for conscious require much more time to process an answer
processing. The importance of this pre-conscious consciously. For more complex questions, such
processing in the learning process is only just as clarifying ambiguity, elucidating thinking
beginning to be realised. To illustrate the processes and analysing group-working skills,
significant powers of these pre-conscious routines the teacher needs to employ questioning
read the following text: strategies that use long response times, for
example talk partners, no hands up, etc.
I cdnuolt blveiee taht I cluod aulaclty uesdnatnrd
waht I was rdanieg. The phaonmneal pweor of Baddeley’s amended model has the inner
the hmuan mnid. Aoccdrnig to rscheearch at voice (auditory), mind’s eye (visual) and spatial
Cmabrigde Uinervtisy, it deosn’t mttaer in waht (kinaesthetic) stores and the episodic buffer,
oredr the ltteers in a wrod are, the olny iprmoatnt all linked from working memory through long-
tihng is taht the frist and lsat ltteer be in the rghit term working memory to long-term memory
pclae. The rset can be a taotl mses and you can information.
sitll raed it wouthit a porbelm. Tihs is bcuseae the
huamn mnid deos not raed ervey lteter by istlef,
but the wrod as a wlohe. Amzanig huh? Yaeh
and I awlyas thought slpeling was ipmorantt.

SSR September 2010, 92(338) 109

Working memory and learning Crossland

l be familiar with the successes and problems

Message 15. Working memory can be enhanced
pupils have while engaging with the task; (9, 10,
in terms of its capacity and processing power
12, 13, 14)
by using task-specific stored representations
l devise the beginnings of tasks that fully
in long-term working memory. Tasks that have
engage the VAK aspects of working memory; (15)
components for visual, auditory and kinaesthetic
l devise an increase in challenge part-way
(VAK) stimuli engage the full capacity of working
through the task so that all the individual working
memory and long-term working memory and
memories in the group are stretched beyond their
as a result the pace of learning and subsequent
storage and processing capacities, producing
recall should be more successful.
cognitive conflict but not to the point where the
pupils ‘switch off’ the task. Challenge must cause
cognitive conflict so that the brain has to adapt to
Conclusions the new demands by rewiring itself. (3, 9, 11).
In the 21st century, we are discovering more and Organising a teacher-led introduction
more about the brain and the role of emotion, and The teachers found that using an acronym helped
challenging old ideas about how we learn, make them to think through the decision-making
decisions, act and remember. (Frith 2008) about the introduction. The acronym GILS
arose out of the metaphor of ‘pupils drowning
The purpose of this article and the previous one in the complications of the task’. During the
(Crossland, 2010) is to report how a group of introduction, the teacher provides GILS to enable
ASTs moved this discussion into the classroom the pupils to engage with the task:
by testing the implications from research models
against their own outstanding practice. The G Group-working skills. Explicitly teach about
individual messages for teachers cannot form the emotional context (e.g. using the GTS
a coherent psychological learning theory or criteria from Box 1 in the previous article
pedagogy because the current understanding of (Crossland, 2010)) so that during the task
the way the brain functions is still in its infancy. the group members can monitor their own
However, the messages give credence to best and the group’s performance against a shared
practice by forming a useful checklist for aligning set of criteria. The explicit teaching of these
learning strategies more closely to the natural skills was the most unexpected and successful
way the human brain learns. Further, grouping the outcome for improving learning arising out of
messages together into sets provides very useful the North Yorkshire research. (2, 4, 5, 8)
guidance for teachers when: I Instruction. Provide the purpose of the task
as the last instruction. If the instruction is
l devising and planning challenging tasks; longer than two seconds then the teacher needs
l organising a teacher-led introduction; to plan a strategy so that all the pupils can
l supporting collaborative group work; remember the ultimate purpose of the task.
l leading a learning review/plenary. For example, repeat the purpose several times
For the following conclusions, the numbers in the during the introduction, ask individual pupils
brackets refer to the related messages in this and to repeat it or talk to other pupils about it or
the previous article (Crossland, 2010). make up a short visual representation. (12)
L Language. Provide the definition of new
Devising and planning challenging tasks words and clarify the meaning of difficult
The teacher needs to: words related to the task so that the pupils’
l plan a classroom setting that promotes an inner voices can immediately engage with
emotionally positive atmosphere; (2, 4, 5) talking about the task. (9, 13)
l have a secure knowledge of the range of S Skills. Ensure that the pupils have the basic
prior learning for the class, that is the range of manipulative and computational skills to
pre-conscious representations; (14, 15) engage with the task. (4, 9, 15)
l understand the conceptual structure of the A word of caution when organising the
curriculum material related to the task; (4) introduction is to be wary about bridging
from previous learning to a new task as it can
110 SSR September 2010, 92(338)
Crossland Working memory and learning

sometimes make the introduction too long; longer classroom by introducing something that they
than five minutes can lead to a drop in the overall currently do not do. Many of the messages and
pace of learning for the lesson. If bridging occurs notes are difficult to implement without additional
in the introduction, the new learning still needs to support from an external trainer, head teacher or
be bridged back into previous learning during the teaching colleagues, and, equally important, access
review/plenary. (6, 7) to exemplar curriculum materials such as Active
assessment (Naylor, Keogh and Goldsworthy,
Supporting collaborative group work
2004), Let’s think! (Adey, Robertson and Venville,
The teacher needs to:
2001) and Thinking science (Adey, Shayer and
l stand back (non-intervention) from the groups Yates, 2001). It is only with this level of support
unless a group becomes dysfunctional. For a that most teachers can practise and learn the new
dysfunctional group, the teacher could model open- teaching techniques. The process often takes several
questioning by asking members of the group to say weeks of perseverance to show significant and
how they are working together, as well as describing long-term gains in the pace of learning because the
their understanding about the purpose of the task. For pupils also need time to learn how to make the most
example, instead of answering a question from one of the new learning opportunities. Primary teachers
person about being ‘stuck’, the teacher asks the group have reported that the gains in learning spread
to discuss together the parts of the task they know out from science to become embedded across the
how to do and the parts they are ‘stuck’ on. The curriculum. Secondary teachers report that the new
ensuing discussion usually resolves the problems the teaching techniques apply to all key stages and are
group is having and they begin to make progress in a in line with the recent emphasis on assessment for
collaborative way. With a group that is really unsure, learning and increasing personalised learning.
the teacher could model the mediation of answers
from different members of the group and then Future research
explain that this is a process they can use themselves Recently, Goleman (2008) released a clarification
next time they are having difficulties. (3, 8, 9, 11) about the role of IQ and emotional intelligence:
Leading a learning review/plenary ‘Compared to IQ and expertise, emotional
The teacher needs to: competence mattered twice as much in what
set stars apart from the average. This held true
l ask some groups to present their learning
across all categories of jobs, and in all kinds of
outcomes; through the process of metacognition,
organizations.’ The big-picture model of the brain
pupils learn to regulate their own behaviours in order
showed the importance of the emotional dimension
to improve their learning strategies such as predicting
in learning; this was supported by the positive
outcomes, planning ahead, apportioning time more
impact of explicitly teaching group-working skills.
effectively, explaining current understanding and
Teachers need to be involved in further action
misunderstandings, and understanding the part–
research into the characteristics of emotional
whole relationships of the components in the overall
development at different ages. This could lead to
structure of the task. Metacognition creates self-
drawing up implications for teaching the skills
awareness in learners so they can sustain their own
required to work in groups at different ages, such as
learning and become lifelong learners. (6, 7, 14)
the what, when and how of introducing group roles.
l ask bridging questions to reinforce new learning
Our experience showed that key stage 2 pupils
where individual pupils link new learning to their
(ages 7–11) were prepared to engage in playing
previous learning in order to move towards a more
a prescribed role of a presenter, that is the person
expert performance. The learning could have
who presents the group outcomes, but it was not
occurred in previous lessons from science or other
until key stage 3 (ages 11–14) that the group could
subjects, particularly maths and English, or from
organise themselves to produce a group presentation
everyday experiences. As each person has a unique
with smooth transitions between presenters.
set of memories, this is not a teacher or group
response but an individual pupil response. (6, 7, 14) Acknowledgements

Our experience of using the messages and My thanks to Simon Carson, Heather Marsh and
concluding notes during in-service courses is that Sarah Jones who all currently teach in North
most teachers can enhance the learning in their Yorkshire schools.

SSR September 2010, 92(338) 111

Working memory and learning Crossland

References
Adey, P., Robertson, A. and Venville, G. (2001) Let’s think! controlled trial. Journal of the American Academy of
(CASE programme designed to develop the thinking Child & Adolescent Psychiatry, 44(2), 177–186.
abilities of 5–6 year-olds). London: nferNelson. Mercer, N. (2000) Words and minds: how we use language
Adey, P., Shayer, M. and Yates, C. 3rd edn (2001) Thinking to think together. Clevedon: Multilingual Matters.
science – the materials of the CASE project (programme Meulen M. (2008) Exploring the interaction between
for KS3). Cheltenham, Gloucestershire: Nelson Thornes. working memory and long-term memory: evidence
Baddeley, A. D. (2000) The episodic buffer: a new for the workspace model. PhD Thesis, University of
component of working memory? Trends in Cognitive Edinburgh.
Sciences, 4, 417–423. Naylor S., Keogh B. with Goldsworthy, A (2004) Active
Cowan, N. (2001) The magical number 4 in short-term Assessment: thinking, learning and assessment in science.
memory: a reconsideration of the mental storage capacity. Sandbach, Cheshire: Millgate House.
Behavioural and Brain Sciences, 23, 87–185. Oberauer, K. (2002) Access to information in working
Crossland, J. (2010) Brain biology and learning. School memory: exploring the focus of attention. Journal of
Science Review, 91(337), 99–107. Experimental Psychology: Learning, Memory, and
Frith C. (2008) No one really uses reason. New Scientist, Cognition, 28, 411–-421.
26 July, 45. Shayer, M. and Adhami, M. (2007) Fostering cognitive
Goleman. D. (2008) www.danielgoleman.info/blog/topics/ development through the context of mathematics:
emotional-intelligence/clarifications (accessed April results of the CAME project. Educational Studies in
2009). Mathematics, 64(3), 265–291.
Klingberg, T., Fernell, E., Olesen, P., Johnson, M., Vygotsky, L. (1978) Interaction between learning and
Gustafsson, P., Dahlstrom, K., Gillberg, C., Forssberg, development. In Mind in society, trans. Cole, M. pp.
H. and Westerberg, H. (2005) Computerized training of 79–91. Cambridge, MA: Harvard University Press.
working memory in children with ADHD – a randomized,

John Crossland is the former senior adviser for science in North Yorkshire and is now an independent
educational consultant. Email: [email protected]; website: www.johncrossland.com

Host a scientist and enrich classroom teaching

Our enthusiastic research scientists are offering you
14 hours of student contact time to:
• enrich learning with curriculum-relevant projects
• engage and inspire students whilst improving grades
• expose students to exciting career options
• help maintain your CPD

Apply now to host a research scientist in your school:

www.researchersinresidence.ac.uk or call 0845 365 7470

Researchers in Residence is free to UK secondary schools and is funded by

Research Councils UK (RCUK) with support from the Wellcome Trust.

112 SSR September 2010, 92(338)

View publication stats