You can listen to a podcast while walking down a busy street. You cannot listen to two podcasts at once. Something in there can run two jobs in parallel and something else cannot — and working out which is which is how this model was built.
📚 What you need to know
The working memory model (WMM) was devised by Baddeley & Hitch (1974) as a more dynamic and flexible model of memory.
It covers short-term memory only. It explains the ‘here and now’ and makes no provision for LTM.
It treats STM as a complex information processor, not a static unitary store.
Four units: the central executive (CE), the phonological loop (PL), the visuospatial sketchpad (VSS), and the episodic buffer (EB), added much later in 2000.
Cognitive load theory (CLT) says working memory has a limited capability for processing new information, so learning suffers when it is overloaded.
Dual-task studies test this by having people do two things at once to see whether they compete for resources.
The study you need is Baddeley et al. (1973).
The four units
The episodic buffer was a late addition, acknowledging that working memory has to talk to long-term memory to function at all.
The central executive
The CE is the driving force. It decides which tasks go to which slave system, rather like the managing director of a company. It is not really a memory store but a command centre, controlling the processes of working memory: what to attend to, what to ignore, and where to send it. It is probably the most important element in the model and by far the hardest to test.
The phonological loop
The PL handles acoustic information, whether spoken or written. Written words are encoded by vocalising them in the mind, turning visual information into acoustic information. It splits into two parts. The phonological store is the ‘inner ear’, holding speech-based information for a couple of seconds. The articulatory control system translates written words into speech so they can be kept in a loop — which is exactly what you are doing when you repeat a phone number until you can type it in.
The visuospatial sketchpad
The VSS handles visual and spatial information: the colour of a flower, where the windows are in a house, how to cross a crowded room. It also splits in two. The visual cache is the ‘inner eye’, storing form and colour. The inner scribe holds spatial and movement information, and rehearses it into the visual cache.
The episodic buffer
The EB was added in 2000 to acknowledge that the CE and slave systems have to communicate with LTM to work properly. Working memory has to respond to the current situation — not remembering which side of the road to drive on would be a serious problem. The EB arranges information into ‘packets’ and moves them to the slave systems in sequence, so events feel continuous. When you have a conversation, the other person’s speech and their lip movements arrive as one experience rather than two out-of-joint streams.
Cognitive load
Cognitive load theory proposes that working memory has a limited capability for processing new information. When it is overloaded, learning and task performance suffer, because there is not enough capacity to handle competing information at once. The load is simply too heavy.
Cognitive load shows up clearly in dual-task studies, where participants do two things simultaneously to see how the tasks compete for resources — drawing while listening to music, or drawing while trying to visualise shapes. This is the standard way to assess divided attention.
Key study: Baddeley et al. (1973)
Part of the study
What happened
Aim
To investigate whether the VSS and the PL are separate components of working memory, using a dual-task activity.
Procedure
A dual-task lab experiment. Participants had a tracking task — following a spot of light with a pointer around a circular path — while imagining block capital letters such as H, T, F and E in their heads.
Condition A
Participants started at the bottom left of the path and tracked the light while giving verbal yes/no responses about features of an imagined capital letter, for example whether it has a top, middle or bottom line. Thinking of ‘F’, they would answer yes to a top and a middle line and no to a bottom line.
Condition B
Participants tracked the light and were told to imagine a letter, then asked whether it had a top, middle or bottom line — but had to silently visualise the letter’s features while tracking.
Results
Participants in condition B had great difficulty doing both at once. Tracking and letter imagery were competing for the limited resources of the VSS. In condition A the tracking and the verbal responses used separate components — the VSS and the PL — so both could run.
Conclusion
There appear to be separate slave systems operating within working memory. Overloading one of them with two tasks means it cannot function properly.
The logic in one line. Two tasks that use different subsystems can run together. Two tasks that use the same subsystem cannot. That difference is the evidence that the subsystems are actually separate.
Do not just say condition B was harder. Say why: both jobs were visual, so both were queuing for the same sketchpad. That sentence is the difference between describing and explaining.
Evaluating the model
What is strong about it
Dual-task research is replicable, since most studies are lab-based and use standardised procedures that can be repeated to check for consistency.
It captures the dynamic nature of working memory, showing how memory is used in the moment to complete a complex task rather than sitting still.
What is weaker about it
Because participants visualised letters in their heads, there is no way to be sure they were not distracted or thinking about something else. That weakens validity.
Extraneous variables may confound dual-task research. Some people naturally have better memories or longer concentration spans; some are more visual or more auditory learners; some are dyslexic or dyspraxic. Findings may therefore not be due to cognitive load at all, and the research may not account for individual differences.
Link to concepts
Perspective: the WMM gives a far more detailed and dynamic account of STM than the MSM. It explains how responses to acoustic and visual information can happen at the same time, rather than treating memory as linear and static. Most researchers now say ‘working memory’ rather than ‘short-term memory’, because the term captures the here-and-now nature of a system working hard within restricted duration and capacity.
Causality: there is very little evidence for how the CE actually directs attention. It is almost impossible to test, so it remains hypothetical, while the VSS and PL can be measured. The model also fails to explain the role of LTM properly, so it ignores key factors in how LTM and the EB combine.
EXAM PRACTICE
Discuss cognitive load theory with reference to one study. [22]
Step 1: define cognitive load
Working memory has limited capacity for new information; overload impairs learning and performance.
Step 2: locate it in the WMM
Baddeley & Hitch (1974): CE plus PL, VSS and later the EB. Each subsystem has its own limited resources.
Step 3: the study
Baddeley et al. (1973). Tracking a light while working with imagined block capitals.
Step 4: the result and the reasonCondition B was very difficult because tracking and visualising both used the VSS. Condition A worked because the tasks used the VSS and the PL separately.
Step 5: evaluate
Replicable and dynamic, but internal visualising cannot be verified, extraneous variables abound, and the CE remains untestable.
Load + model + study + mechanism + evaluationName which subsystem each task used. Vague answers lose the whole argument.
💡 Exam tip
Learn the sub-parts: phonological store and articulatory control system; visual cache and inner scribe.
Say the WMM covers STM only. It is not a full model of memory.
The untestable CE is your strongest theoretical criticism.
The 2000 date for the episodic buffer is a precise detail worth including.
Dual-task methodology is the design point — explain what it is designed to reveal.
⚠ Common mix-up
Saying the WMM replaced the MSM. It refines one part of it; it does not cover LTM.
Calling the CE a store. It is a control system, not a place where things are kept.
Getting the conditions backwards. Condition B was the difficult one, because both tasks were visual.
Saying the episodic buffer was in the original model. It arrived 26 years later.
Treating ‘slave system’ as an insult. It is simply the model’s term for a subsystem taking orders from the CE.
Up next: Multi-Store or Working Memory: Which Fits Better? — the two models side by side, which is exactly how comparison questions are marked.
Want this explained one-to-one?
Book a free session with an experienced IB Psychology tutor and get your trickiest topics made simple.
