Working Memory vs Short-Term Memory: Key Differences Explained

Working Memory vs Short-Term Memory: What's the Real Difference?

Working memory and short-term memory are often used interchangeably — even by professionals who should know better. But they're not the same thing. Understanding the difference matters because it changes how you think about cognitive training, testing, and everyday brain performance.

This article explains the distinction clearly, with concrete examples, and shows you how CortexLab tests both.

The Simple Distinction

• Short-term memory: The ability to hold information temporarily (seconds to about a minute)
• Working memory: The ability to hold AND manipulate information at the same time

Think of it this way: short-term memory is a notepad. Working memory is a notepad plus a calculator.

Example: Phone Numbers

• Short-term memory: Someone tells you a phone number and you repeat it back. You're just holding the digits
• Working memory: Someone tells you a phone number and asks you to say it backward. Now you're holding the digits and manipulating their order

Example: Mental Arithmetic

• Short-term memory: Remember that the bill is $47.50
• Working memory: Calculate a 20% tip on $47.50 in your head. You need to hold $47.50, multiply by 0.2, and keep track of intermediate steps — all without writing anything down

The Science Behind the Distinction

Short-Term Memory: The Atkinson-Shiffrin Model (1968)

The original model of memory proposed three stages: sensory memory, short-term memory, and long-term memory. In this model, short-term memory was a simple buffer — a temporary holding area with limited capacity (7±2 items) and duration (15–30 seconds without rehearsal).

Working Memory: Baddeley's Model (1974/2000)

Alan Baddeley and Graham Hitch proposed that the simple "short-term store" should be replaced with a more complex system — working memory — with multiple components:

• Central Executive: The control system that directs attention, coordinates information, and manages the other components
• Phonological Loop: Handles verbal and acoustic information (inner voice/inner ear)
• Visuospatial Sketchpad: Handles visual and spatial information (mental imagery)
• Episodic Buffer: Integrates information from the other components and links to long-term memory

For a deeper dive into this model, see our article on Baddeley's working memory model.

Key Differences Summarized

Feature	Short-Term Memory	Working Memory
Function	Passive storage	Active storage + processing
Capacity	7±2 items (Miller)	4±1 chunks (Cowan)
Components	Single system	Multiple subsystems
Cognitive load	Low	High (manipulation required)
Brain regions	Primarily temporal lobe	Prefrontal cortex + parietal + temporal
Predicts	Recall accuracy	IQ, academic performance, job performance

Why the Difference Matters

For Cognitive Testing

Tests that only measure short-term memory (digit span forward, pattern recall) tell you about storage capacity. Tests that measure working memory (digit span backward, N-back, operation span) tell you about cognitive processing power — which is far more predictive of real-world performance.

For Training

If you want to improve cognitive performance, training working memory (with manipulation demands) is more effective than training short-term memory (with storage-only demands). This is why N-back tasks — which require continuous updating and comparison — are the most studied working memory training method.

For Understanding Cognitive Difficulties

Many conditions affect working memory more than short-term memory:

• ADHD: Short-term storage may be relatively intact, but the working memory manipulation component is significantly affected
• Aging: Simple span tasks decline slowly; complex span tasks (working memory) decline faster
• Stress: Acute stress impairs the central executive (working memory control) more than passive storage
• Sleep deprivation: Working memory is more sensitive to sleep loss than short-term memory

How CortexLab Tests Both

Memory Grid — Visuospatial Short-Term / Working Memory

The Memory Grid test shows you a pattern of highlighted cells, then asks you to reproduce it. This primarily tests visuospatial short-term memory (the visuospatial sketchpad), but at higher difficulty levels (4×4 and 5×5 grids), working memory is increasingly engaged as you need to hold and organize more complex spatial information.

• Levels 1–2 (3×3, small 4×4): Primarily short-term memory — can you store the pattern?
• Levels 3–5 (larger patterns): Working memory — can you organize and reconstruct complex spatial information?

DSST — Processing Speed + Working Memory

The Digit Symbol Substitution Test requires you to hold the digit-symbol coding key in working memory while rapidly matching pairs. It's a working memory task because you must simultaneously store the key and apply it — storage plus manipulation.

Task Switching — Executive Working Memory

Task Switching tests the central executive — the "boss" component of working memory. Switching between two rules requires updating the active rule set in working memory while suppressing the previous one.

Practical Implications

If You Want to Improve Short-Term Memory

• Practice chunking: Group information into meaningful units
• Use rehearsal: Repeat information to prevent decay
• Leverage multiple modalities: Combine visual and verbal encoding

If You Want to Improve Working Memory

• N-back training: The most evidence-based approach
• Aerobic exercise: Strengthens prefrontal cortex function
• Sleep optimization: Working memory is highly sensitive to sleep quality
• Reduce cognitive load: Use external tools (notes, checklists) to free up working memory for the tasks that need it

Short-term memory holds information; working memory holds it and works with it. This distinction isn't just academic — it determines which tests give you meaningful data, which training methods actually help, and how to interpret cognitive difficulties. Use CortexLab's free tests to measure both your storage capacity and processing power, then target your training where it matters most.