The Multi-Store Model of Memory

The multi-store model (MSM) was proposed by Atkinson and Shiffrin (1968). It is one of the most influential models of memory in psychology and describes memory as a linear process flowing through three distinct stores: the sensory register, short-term memory (STM), and long-term memory (LTM).

---

Structure of the Multi-Store Model

The MSM proposes that information passes through memory in a fixed sequence:

flowchart LR
    A[Sensory Input] --> B["Sensory Register<br/>iconic / echoic<br/>< 4 sec"]
    B -->|Attention| C["Short-Term Memory<br/>7 plus or minus 2<br/>18-30 sec"]
    C -->|Maintenance<br/>rehearsal| D["Long-Term Memory<br/>unlimited capacity<br/>up to a lifetime"]
    B -.->|No attention| E[Decay / lost]
    C -.->|Decay or<br/>displacement| E
    D -->|Retrieval| C

Each store differs in terms of encoding, capacity, and duration:

Feature	Sensory Register	Short-Term Memory	Long-Term Memory
Encoding	Modality-specific (visual, auditory, etc.)	Mainly acoustic	Mainly semantic
Capacity	Very large (all sensory input)	Limited (7 ± 2 items)	Potentially unlimited
Duration	Very brief (< 1 second visual, 2–4 seconds auditory)	18–30 seconds without rehearsal	Potentially a lifetime

---

The Sensory Register

The sensory register is the first store in the MSM. It receives information from the senses (sight, hearing, touch, taste, smell). There is a separate sensory store for each sense:

• Iconic memory — visual sensory store (lasting about 0.5 seconds)
• Echoic memory — auditory sensory store (lasting about 2–4 seconds)

Information in the sensory register is stored in its raw, unprocessed form. Most of this information decays rapidly unless attention is paid to it, in which case it is transferred to STM.

---

Short-Term Memory (STM)

STM holds a limited amount of information for a short duration. According to the MSM:

• Information enters STM from the sensory register when we pay attention to it
• STM encodes information primarily acoustically (by sound)
• STM has a capacity of approximately 7 ± 2 items (Miller, 1956)
• Information is lost from STM through decay (fading over time) or displacement (being pushed out by new information)
• If information is rehearsed (repeated), it can be transferred to LTM

The process of repeating information to keep it in STM or to transfer it to LTM is called maintenance rehearsal. Atkinson and Shiffrin argued that the more information is rehearsed, the more likely it is to be transferred to LTM.

---

Long-Term Memory (LTM)

LTM is the permanent memory store. According to the MSM:

• Information enters LTM through prolonged rehearsal in STM
• LTM encodes information primarily semantically (by meaning)
• LTM has a potentially unlimited capacity
• LTM has a potentially unlimited duration — memories can last a lifetime
• Information is retrieved from LTM back into STM when needed

---

Evidence Supporting the MSM

The Case of HM (Henry Molaison)

One of the most famous case studies in memory research involved a patient known as HM. In 1953, HM had brain surgery to treat severe epilepsy, which removed parts of his hippocampus. After the surgery:

• His STM was intact — he could hold conversations and remember small amounts of information briefly
• His LTM was severely impaired — he could not form new long-term memories (a condition called anterograde amnesia)
• His existing long-term memories from before the surgery were largely preserved

This case supports the MSM because it shows that STM and LTM are separate stores — damage to one does not necessarily affect the other.

The Serial Position Effect

Glanzer and Cunitz (1966) asked participants to remember a list of words. They found:

• Words at the beginning of the list were recalled well — the primacy effect (these words had been rehearsed and transferred to LTM)
• Words at the end of the list were recalled well — the recency effect (these words were still in STM)
• Words in the middle were recalled poorly

This supports the MSM because it suggests that two separate stores (STM and LTM) are involved in recall.

---

Evaluation of the Multi-Store Model

Strengths

• It is supported by research evidence (e.g. HM, serial position effect, Peterson and Peterson)
• It was one of the first models to describe memory as a process with distinct stages
• The distinction between STM and LTM is well established in cognitive psychology

Weaknesses

• It is too simplistic — it treats STM and LTM as single, unitary stores, when research suggests there are different types of each (e.g. episodic, semantic, and procedural LTM)
• Maintenance rehearsal is not the only way information enters LTM — we can remember things without deliberately rehearsing them (e.g. flashbulb memories)
• The model is linear and inflexible — it does not account for how different types of information are processed differently
• The working memory model (Baddeley and Hitch, 1974) provides a more detailed account of STM

Exam Tip: When evaluating the MSM, always use specific research evidence to support your points. For example, you could use HM as evidence that STM and LTM are separate, and then note that the model is too simplistic because LTM has been shown to consist of multiple types.

---

Key Points

• The MSM was proposed by Atkinson and Shiffrin (1968).
• Memory flows through three stores: sensory register → STM → LTM.
• Each store differs in encoding, capacity, and duration.
• Maintenance rehearsal transfers information from STM to LTM.
• The model is supported by case studies (HM) and the serial position effect.
• The model is criticised for being too simplistic and linear.

---

Study case: The case of HM (Henry Molaison)

Aim: To understand the contribution of the hippocampus and surrounding medial temporal lobe structures to memory formation, following major brain surgery. The case also provided a unique natural test of the multi-store model's claim that STM and LTM are separate stores.