Working memory in children: What parents and teachers need to know


Working memory is often likened to RAM in a computer. The more you have, the more information you can juggle at once — allowing for faster processing. But of course we humans can’t improve our memory capacity by installing a new RAM stick. Nor can we expect new humans to perform at the highest level “straight out of the box.” Our skills grow over the course of childhood.

So what are the different components or subsystems of working memory? How do they develop? And what can we do to help children who are struggling? Here is an evidence-based guide.

little boy smiling and peaking out from behind book

What is working memory, and why is it important?

Working memory, or WM, is a bundle of mechanisms that allows us to track information in real time. It’s what we use to plan and carry out an action — the mental workspace where we manipulate information, crunch numbers, or see with our “mind’s eye” (Cowan 2010; Miller et al 1960).

I’ve seen many definitions, but two strike me as particularly helpful. The first is from Nelson Cowan, a world expert in the development of working memory and attention:

“Working memory is the retention of a small amount of information in a readily accessible form” (Cowan 2014).

And the second is from cognitive scientist Bradley Buchsbaum:

“Working memory is a cognitive system for the maintenance, manipulation, and monitoring of information that is not currently available in the sensory environment” (Buchsbaum 2013).

In other words, WM is the reason you don’t immediately forget what someone just told you, or what you saw just a moment before. It’s the reason you remember what comes next when you are working through a multi-step task.

What are some everyday examples of working memory?

Here are several — each illustrating a different component, or subsystem, of working memory.

Example: The phonological loop

Let’s say you provide verbal instructions to your child — a list of things to put in his backpack. “Please pack your lunch, your medication, your homework, and that permission slip I signed.”

Your child hears what you say, and follows through — in part because he keeps “replaying” your verbal instructions in his mind before the memory fades away. This type of verbal working memory is called the “phonological loop.”

Example: The visuospatial sketchpad

Imagine that your child takes a brief peak through a window at a man and a dog. Immediately after, she can still conjure up the image she saw. It’s temporarily “cached” in the component of WM called the “visuospatial sketchpad,” and she can consult that sketchpad to answer questions. What kind of dog was it? What was the man wearing? Which direction were they walking?

Example: The central executive

Suppose you are trying to do some calculations in your head — figuring out the final price of a book after applying a discount of 15%. You might use your visuospatial sketchpad to “see” the number-crunching in your “mind’s eye.” But you also need to keep updating the imagery as you multiply and add. So you need something to supervise the process, and this component of the working memory system is called the “central executive.”

Example: The episodic buffer

Alan Baddeley has proposed a fourth component of WM called the “episodic buffer.” It’s what allows us to link together different types of information (like sights, sounds, and spatial coordinates) into one, coherent episode (Baddeley 2000).

An example of this would be meeting a new person. You hear the person’s name, see the person’s face, and observe facial expressions and body language. Tracking this simulataneously depends on the episodic buffer (Kofler et al 2018).

What is working memory capacity?

Working memory capacity, or WMC, refers to how many items you can track at once. It isn’t very much, even if you happen to be a working memory whiz. In normal adults, the capacity limit typically ranges between 3 and 5 items (Cowan 2010).

But, as you might guess, this individual variation is meaningful. People with larger capacities can juggle more information. This helps them process information more quickly, and the benefits are well-documented. People with higher-than-average WMC are more likely to excel in the classroom.

For example, when researchers have tracked the development of primary school children, they’ve found that early gains in WM — especially in visuospatial memory — predicted later achievement in mathematics (Li and Geary 2013; Li and Geary 2017).

Working memory is also predictive of language skills, like the ability to keep track of the ideas presented in a long or complex sentence  (Zhou et al 2017).

On the flip side, individuals with poor WM skills at a disadvantage. They are more likely to struggle with mathematics and reading. They may also struggle with following spoken directions. There is too much to juggle, and they lose track of what they are supposed to do.

But what’s developmentally normal? Doesn’t WM improve as children get older?

Yes. When researchers have administered the same WM tests across different developmental stages, they’ve found evidence for steady improvement, with adults performing almost twice as well as young children (Gatherole et al 2004; Gatherole and Alloway 2007).

For example, in WM tasks dependent on tracking items in a briefly-presented visual array, adults remembered approximately 3 or 4 objects (Cowan 2016). Five-year-olds recalled only half as many (Riggs et al 2006).

Overall, WM performance shows a slow-and-steady rate of improvement from the preschool years to the age of 15 (Cowan 2016).

So how can we tell if a child has a low WM capacity for his or her age?

Researchers estimate that 10-15% of school age children are struggling with low working memory capacity (Holmes et al 2009; Holmes et al 2010; Fried et al 2016). How can we identify these kids?

A professional diagnosis depends on administering special tests, like the Comprehensive Assessment Battery for Children – Working Memory (CABC-WM), or the Automated Working Memory Assessment (which you can read about here).

But you can also look for everyday signs. According to Susan Gatherole and Tracey Alloway (2007), kids with working memory difficulties typically

  • have normal social relationships with peers;
  • are reserved during group activities in the classroom, and sometimes fail to answer direct questions;
  • find it difficult to follow instructions;
  • lose track during complicated tasks, and may eventually abandon these tasks;
  • make place-keeping errors (skipping or repeating steps);
  • show incomplete recall;
  • appear to be easily distracted, inattentive, or “zoned out”; and
  • have trouble with activities that require both storage (remembering) and processing (manipulating information).

Do poor WM skills mean that a child isn’t smart? Do strong WM skills mean that a child is highly intelligent?

No. Working memory contributes to intelligence. It affects how we learn. It helps us stay focused when there are distractions. It can have an impact on how well we perform on tests, including achievement tests and IQ tests. But we can’t equate WM with overall intelligence. 

For instance, take “fluid intelligence” — what psychologists define as “the ability to reason through and solve novel problems” (Shipstead et al 2016).

When researchers administered tests of fluid intelligence and working memory to young school children (aged 5-7 years), they found a correlation: Kids who performed well on one test tended to perform well on the other (Ger and Roebers 2023).

However, only one type of working memory (verbal WM) was correlated with fluid intelligence, and the correlation itself was pretty small.

And if we think about what skills help us solve new problems, we can see why this might be.

Fluid intelligence doesn’t just demand that we keep relevant information in mind. It also requires that we discard — stop thinking about — information that is irrelevant. We need to forget outdated ideas in order to make room for new ones (Shipstead et al 2016). 

Thus, it isn’t so much the size of mental notepad that matters, but whether we are filling that notepad with the most promising information. Merely having a larger WM capacity doesn’t necessarily enhance your fluid intelligence.

Then there is the evidence from IQ tests: Working memory capacity doesn’t always correlate with IQ.

Some kids perform well on IQ tests and yet have relatively mediocre WM skills (Alloway and Alloway 2010). How is this possible? Tests like the Wechsler Intelligence Scale for Children (WISC) have distinct subtests. Some specifically target working memory. Others don’t.

In addition, there are components of intelligence that go largely unmeasured by IQ tests, and don’t correlate with working memory capacity.

One example is rationality and logic. It’s a reflective mode of thought that IQ tests ignore. But it’s essential for making smart decisions, and it’s not clear that working memory capacity has much of an impact. In recent experiments, people with higher WMCs were just as likely as other folks to experience biased, faulty reasoning (Robinson and Unsworth 2017).

Finally, it’s important to remember that working memory isn’t a single, unitary system.

As noted above, there are different types of WM, and each type is associated with different types of thinking.

For instance, in the study of children’s fluid intelligence mentioned above, verbal working memory was linked with higher fluid intelligence, but visual-spatial WM wasn’t (Ger and Roebers 2023).

By contrast, both verbal and visual-spatial WM are linked with better performance in mathematics (Zhang et al 2022; Liang at al 2022; Fanari et al 2019). So is “series order” working memory — the ability to keep track of sequences (Attout and Majerus 2018; Carpenter et al 2018).

What about learning disabilities and developmental disorders?

Once again, there are links. For example, when kids suffer from dyscalculia (a mathematical learning disability), they tend to show weaker visuaol-spatial and “series order” WM skills (Attout and Majerus 2015: Menon 2016). 

Working memory problems can also make it more difficult for young children learning to read. Deficits in verbal WM have been linked with reading comprehension problems in older children (Peng et al 2018). 

In addition, kids with autism are also more likely to experience working memory problems, with deficits in spatial WM being more common than deficits in verbal working memory (Wang et al 2017). And children with attention deficit hyperactivity disorder (ADHD) are more likely than normally-developing kids to suffer from impairments of verbal working memory (Ramos et al 2019; Kennedy et al 2019).

Can we improve WM performance by playing simple memory games?

happy, primary school children in classroom - girl on laptop with boy watching

You might have heard of computer-based memory games that are supposed to enhance WM, or even IQ. Do they actually work? It depends on what you mean by “work.”

For example, consider the computer-based training program developed by Cogmed.

In one study, researchers identified kids with low WMC, and assigned these children to play a series of computer games designed to challenge their WM skills (Holmes et al 2009). Some of these games  included:

  • Hearing a series of letters read aloud (“G, W, Q, T, F…”) and repeating them back.
  • Watching a battery of lamps light up, one at a time, and then recalling the correct sequence by clicking the correct locations with a computer mouse.
  • Hearing and watching a sequence of numbers while they are spoken aloud and flashed on a keypad. After each sequence, the student is asked to reproduce the sequence in reverse order by hitting the correct digits on the keypad.

For children in a control group, the difficulty level of these tasks remained easy throughout the study. But for kids in the treatment group, the program was adaptive, i.e., a student was given progressively more difficult tasks as his or her performance improved.

After about 6 weeks of training, researchers re-tested the students’ working memory skills, and the results were pretty dramatic. While both groups improved, the kids in the adaptive program did much better. Their average gains were 3 to 4 times higher than those of kids in the control group.

But there was a crucial catch: Improvements were found only on tests that closely resembled the training games. And that has been the pattern in other studies (Gobet and Sala 2023).

Training helps people get better at the tasks for which they are trained. But the effect doesn’t carry over to other situations.

“Far transfer effects” haven’t panned out — not in the largest, best-designed, most carefully controlled studies conducted to date (Sala and Gobet 2020; Melby-LervÃ¥g et al 2016; Shiphead et al 2012).

So if you are interested in improving a child’s performance on tasks highly similar to the tasks encountered in working memory games, then this type of training is worthwhile. And perhaps someday we’ll find out these games deliver long-term benefits that researchers haven’t yet been able to detect.

But if you’re goal is to enhance a child’s general cognitive abilities, WM training is unlikely to have much impact. Not by itself. A more promising approach is to target the domain-specific skills your child needs to succeed.

For instance, if a child struggles with mathematics, seek out special training in the relevant mathematical skills — like counting, number sense, or basic arithmetic calculations (Kyttälä et al 2015). If a child is having trouble with reading, look for programs that designed for kids who need to build literacy skills (Melby-LervÃ¥g et al 2016).

What else can we do?

As Susan Gathercole and Tracey Alloway note, we can help children compensate for WM limitations in a variety of ways, including these:

  • We can break down tasks into smaller subroutines, so kids can tackle just one component at a time.
  • We can adjust the way we communicate, so we don’t introduce too much material at once, and provide children with regular reminders of what they need to do next.
  • We can ask kids to repeat back new information, and help them connect it with what they already know.
  • We can prompt kids with regular reminders of what to do next, and encourage them to ask questions when they feel lost.
  • We can teach them how to create and use their own memory aids — like taking notes.

And research suggests other tactics too. To get the most from your WMC, you need to understand how it functions. What disrupts WM? What tricks allow people to pack more data in the mental workspace?

For more information, check out these evidence-based tips for improving working memory performance.


References: WM in children

Alderson RM, Kasper LJ, Patros CH, Hudec KL, Tarle SJ, Lea SE. 2015. Working memory deficits in boys with attention deficit/hyperactivity disorder (ADHD): An examination of orthographic coding and episodic buffer processes. Child Neuropsychol. 21(4):509-30.

Alloway TP and Alloway RG. 2010. Investigating the predictive roles of working memory and IQ in academic attainment. Journal of Experimental Child Psychology 106(1): 20-29.

Alloway TP. 2007. Automated working memory assessment. Oxford: Harcourt.

Attout L and Majerus S. 2018. Serial order working memory and numerical ordinal processing share common processes and predict arithmetic abilities. Br J Dev Psychol. 36(2):285-298.

Attout L and Majerus S. 2015. Working memory deficits in developmental dyscalculia: The importance of serial order. Child Neuropsychol. 21(4):432-50.

Baddeley A. 2000. The episodic buffer: a new component of working memory? Trends Cogn Sci. 4(11):417-423.

Buchsbaum BR. 2013. The role of consciousness in the phonological loop: hidden in plain sight. Front Psychol. 4:496.

Carpenter AF, Baud-Bovy G, Georgopoulos AP, Pellizzer G. 2018. Encoding of Serial Order in Working Memory: Neuronal Activity in Motor, Premotor, and Prefrontal Cortex during a Memory Scanning Task. J Neurosci. 38(21):4912-4933.

Cowan N. 2022. Working memory development: A 50-year assessment of research and underlying theories. Cognition. 224:105075. 

Cowan N. 2014. Working Memory Underpins Cognitive Development, Learning, and Education. Educ Psychol Rev. 26(2):197-223.

Cowan N. 2016. Working Memory Maturation: Can We Get at the Essence of Cognitive Growth? Perspect Psychol Sci. 11(2):239-64.

Cowan N. 2010. The Magical Mystery Four: How is Working Memory Capacity Limited, and Why? Curr Dir Psychol Sci. 19(1):51-57.

Cowan N. 2001.The magical number 4 in short-term memory: a reconsideration of mental storage capacity. Behav Brain Sci. 24(1):87-114; discussion 114-85.

Fanari R, Meloni C, Massidda D. 2019. Visual and Spatial Working Memory Abilities Predict Early Math Skills: A Longitudinal Study. Front Psychol. 10:2460.

Fried R, Chan J, Feinberg L, Pope A, Woodworth KY, Faraone SV, Biederman J. 2016. Clinical correlates of working memory deficits in youth with and without ADHD: A controlled study. J Clin Exp Neuropsychol.38(5):487-96.

Gathercole SE, Pickering SJ, Ambridge B, Wearing H. 2004. The structure of working memory from 4 to 15 years of age. Dev Psychol. 40(2):177-90.

Gathercole SE and Alloway TP. 2007. Understanding working memory. London: Harcourt.

Ger E and Roebers CM. 2023. The Relationship between Executive Functions, Working Memory, and Intelligence in Kindergarten Children. J Intell. 11(4):64.

Gobet F and Sala G. 2023. Cognitive Training: A Field in Search of a Phenomenon. Perspect Psychol Sci. 18(1):125-141.

Holmes J, Gathercole SE, and Dunning DL. 2009. Adaptive training leads to sustained enhancement of poor working memory in children. Dev Sci. 12(4):F9-15.

Holmes J, Gathercole S and Dunning D. 2010. Poor working memory. Advances in Child Development and Behavior – ADVAN CHILD DEVELOP BEHAV. 39: 1-43.

Jaeggi, S. M., Buschkuehl, M., Jonides, J., & Perrig, W. J. (2008). Improving Fluid Intelligence With Training on Working Memory. Proceedings of the National Academy of Sciences of the United States of America, 105(19), 6829-6833.

Kennedy RJ, Quinlan DM, Brown TE. 2019. Comparison of Two Measures of Working Memory Impairments in 220 Adolescents and Adults With ADHD. J Atten Disord. 23(14):1838-1843.

Kofler MJ, Spiegel JA, Austin KE, Irwin LN, Soto EF, Sarver DE. 2018. Are Episodic Buffer Processes Intact in ADHD? Experimental Evidence and Linkage with Hyperactive Behavior. J Abnorm Child Psychol. 46(6):1171-1185.

Kuhn JT, Ise E, Raddatz J, Schwenk C, Dobel C. 2016. Basic numerical processing, calculation, and working memory in children with dyscalculia and/or ADHD symptoms. Z Kinder Jugendpsychiatr Psychother. 44(5):365-375.

Kyttälä M, Kanerva K, Kroesbergen E.2015. Training counting skills and working memory in preschool. Scand J Psychol. 56(4):363-70.

Li Y  and Geary DC. 2013. Developmental gains in visuospatial memory predict gains in mathematics achievement. PLoS One. 8(7):e70160.

Li Y and Geary DC. 2017. Children’s visuospatial memory predicts mathematics achievement through early adolescence. PLoS One. 13;12(2):e0172046.

Liang Z, Dong P, Zhou Y, Feng S, and Zhang Q. 2022. Whether verbal and visuospatial working memory play different roles in pupil’s mathematical abilities. Br J Educ Psychol. 92(2):e12454.

Melby-LervÃ¥g M, Redick TS, Hulme C. 2016. Working Memory Training Does Not Improve Performance on Measures of Intelligence or Other Measures of “Far Transfer”: Evidence From a Meta-Analytic Review. Perspect Psychol Sci.11(4):512-34.

Menon V. 2016. Working memory in children’s math learning and its disruption in dyscalculia. Curr. Opin. Behav. Sci. 10:125–132.

Miller GA, Galanter E, Pribram KH. 1960. Plans and the structure of behavior. New York: Holt, Rinehart and Winston.

Peng P, Barnes M, Wang C, Wang W, Li S, Swanson HL, Dardick W, Tao S. A meta-analysis on the relation between reading and working memory. Psychol Bull. 144(1):48-76.

Ramos AA, Hamdan AC, Machado L. 2019. A meta-analysis on verbal working memory in children and adolescents with ADHD. Clin Neuropsychol. 22:1-26.

Riggs KJ, McTaggart J, Simpson A, Freeman RP. 2006. Changes in the capacity of visual working memory in 5- to 10-year-olds. J Exp Child Psychol. 295(1):18-26.

Robison MK and Unsworth N. 2017. Individual differences in working memory capacity and resistance to belief bias in syllogistic reasoning. Q J Exp Psychol (Hove). 70(8):1471-1484.

Sala G and Gobet F. 2020. Working memory training in typically developing children: A multilevel meta-analysis. Psychon Bull Rev. 27(3):423-434.

Sala G and Gobet F. 2017. Working memory training in typically developing children: A meta-analysis of the available evidence. Dev Psychol. 53(4):671-685.

Shah P and Miyake A. 1996. The separability of working memory resources for spatial thinking and language processing: an individual differences approach. J Exp Psychol Gen. 125(1):4-27.

Shipstead Z, Harrison TL, Engle RW. 2016. Working Memory Capacity and Fluid Intelligence: Maintenance and Disengagement. Perspect Psychol Sci. 11(6):771-799.

Shipstead Z, Hicks KL, and Engle RW. 2012. Cogmed working memory training: Does the evidence support the claims? Journal of Applied Research in Memory and Cognition 1 (3): 185–193.

Stanovich KE. 2009. What intelligence tests miss: The psychology of rational thought. Yale University Press.

Tan ASL, Lau RC, Anderson PJ, Gathercole S, Bellgrove MA, Wiley JF, Spencer-Smith MM. 2024. Exploring Working Memory Capacity and Efficiency Processes to Understand Working Memory Training Outcomes in Primary School Children. J Cogn. 7(1):23.

Wang Y, Zhang YB, Liu LL, Cui JF, Wang J, Shum DH, van Amelsvoort T, Chan RC.A Meta-Analysis of Working Memory Impairments in Autism Spectrum Disorders. Neuropsychol Rev. 27(1):46-61.

Zhang Y, Tolmie A, Gordon R. 2022. The Relationship between Working Memory and Arithmetic in Primary School Children: A Meta-Analysis. Brain Sci. 13(1):22.

Zhou H, Rossi S, Chen B. 2017. Effects of Working Memory Capacity and Tasks in Processing L2 Complex Sentence: Evidence from Chinese-English Bilinguals. Front Psychol. 20(8):595.

Content last modified 3/2024

Portions of the text derive from earlier versions of this article, written by the same author.

Title image of boy smiling and peaking from behind book by istock / vipin jaiswal

image of childen looking at laptop by zGel /istock