Atkinson and Shiffrin (1968) proposed that memory consists of three distinct stores. Which sequence correctly describes the flow of information through the multi-store model?
A: Long-term store → short-term store → sensory register
B: Sensory register → short-term store → long-term store
C: Short-term store → sensory register → long-term store
D: Sensory register → long-term store → short-term store
In the multi-store model, environmental input first enters the sensory register, where it is held in a raw, modality-specific form for a very brief period. Attention then selects a subset of this information to pass into the short-term store. Maintenance rehearsal can keep information active in the short-term store; sufficient rehearsal transfers it into the long-term store. The model's assumption of a strictly linear flow was later challenged by evidence that information can move bidirectionally between stores and that long-term memory influences what is encoded.
George Sperling (1960) used the partial report method to investigate the sensory register. What did his results reveal about iconic memory?
A: Iconic memory holds approximately 7 items for up to 30 seconds
B: The visual sensory register holds a large amount of information (~9–12 items) but decays within about 500 milliseconds
C: The sensory register only stores information that has been attended to
D: Iconic memory capacity equals long-term memory capacity
Sperling presented participants with a 3×4 grid of letters for 50 ms. In the whole-report condition, participants recalled only 4–5 items — suggesting a small capacity. But in the partial-report condition, where a tone cued which row to recall after the display vanished, performance was much higher (~9 items estimated), as participants could access any row. However, if the tone was delayed by more than ~300–500 ms, the advantage disappeared. This revealed that the visual sensory register (iconic memory) holds a large, complete snapshot of the visual scene but it decays extremely rapidly — within less than a second.
George Miller's influential 1956 paper proposed a capacity limit for short-term memory. What did he conclude?
A: Short-term memory can hold an unlimited number of items if they are rehearsed
B: Short-term memory capacity is approximately 7 ± 2 chunks of information
C: Short-term memory can hold exactly 4 items; no more
D: Short-term memory capacity varies so widely between individuals that a single number is meaningless
Miller reviewed evidence from a range of tasks — digit span, letter span, judgement of tones — and concluded that human information-processing capacity is constrained to approximately 7 ± 2 chunks across these domains. Crucially, he identified the concept of "chunking": items can be grouped into meaningful units, extending effective capacity. For instance, the string 1-9-4-5-1-9-6-8 might be recalled as three chunks: "1945," "19," "68." Miller's estimate was later revised downward by Cowan (2001) to approximately 4 chunks when rehearsal strategies are controlled.
Peterson and Peterson (1959) presented consonant trigrams (e.g., "KMT") and asked participants to recall them after intervals of 3, 6, 9, 12, 15, or 18 seconds, during which they counted backwards to prevent rehearsal. What did this demonstrate?
A: Short-term memory duration is effectively unlimited if items are meaningful
B: Without rehearsal, information in short-term memory decays rapidly — over approximately 15–18 seconds
C: Forgetting in short-term memory is primarily due to interference, not decay
D: Short-term memory duration is approximately 2 minutes
Recall dropped sharply as the retention interval increased: from around 90% at 3 seconds to around 5% at 18 seconds. Because the counting task prevented rehearsal, this decay could not be attributed to displaced rehearsal — it suggested that short-term memory traces fade within about 15–18 seconds without maintenance rehearsal. This provided key evidence for the multi-store model's claim that STM is a temporary, fragile store distinct from LTM. However, debate continued about whether the forgetting was due to pure trace decay or to interference from counting backward.
Conrad (1964) asked participants to recall sequences of visually presented letters. He found that errors were most likely to involve letters that sounded similar (e.g., confusing B with V or P). What does this reveal about coding in short-term memory?
A: Short-term memory uses predominantly visual/iconic coding
B: Short-term memory uses predominantly semantic (meaning-based) coding
C: Short-term memory uses predominantly acoustic (sound-based) coding, even for visually presented material
D: Short-term memory uses randomly selected coding depending on the stimulus
Even though letters were presented visually, recall errors clustered around acoustically similar letters rather than visually similar ones (e.g., mistaking B for D would be a visual error; mistaking B for V is an acoustic error). This acoustic confusion effect implies that when verbal material enters STM, it is typically recoded into a phonological (sound-based) representation. This contrasted with long-term memory, where Baddeley (1966) demonstrated that semantic similarity — not acoustic similarity — caused interference, suggesting LTM uses semantic coding.
The multi-store model proposes that maintenance rehearsal is the mechanism by which information transfers from STM to LTM. What does research on the levels-of-processing framework suggest about this claim?
A: Maintenance rehearsal is always effective — more rehearsal always produces better long-term retention
B: Maintenance rehearsal alone is not sufficient for durable long-term memory; elaborative, meaning-based processing produces much stronger retention
C: Rehearsal has no effect on memory transfer; only emotional arousal matters
D: Maintenance rehearsal is effective for semantic material but not for phonological material
Craik and Lockhart (1972) proposed that memory strength depends on the depth at which information is processed, not merely the amount of rehearsal. Shallow processing (repeating sounds or letters — maintenance rehearsal) produces weak, fragile traces. Deep, semantic processing (thinking about meaning, forming associations) produces much more durable memories. Experiments showed that simply repeating a word many times produces poor long-term recall compared to processing its meaning once — challenging the multi-store model's assumption that rehearsal equals transfer.
The serial position effect (Murdock, 1962) describes a characteristic pattern when people recall a list of words. How does this pattern support the multi-store model?
A: Words in the middle of a list are recalled best, reflecting both STM and LTM contributions
B: The primacy effect (better recall for early items) is attributed to LTM rehearsal, and the recency effect (better recall for final items) is attributed to STM
C: Both primacy and recency effects reflect LTM storage, disconfirming the STM/LTM distinction
D: The recency effect disappears when participants are given unlimited time to recall items
When people recall a list of words in any order (free recall), they show better memory for words at the beginning (primacy) and end (recency) than for middle items. The primacy effect is attributed to early items receiving more rehearsal opportunities, transferring to LTM. The recency effect reflects items still active in STM at the time of recall. Evidence supporting this interpretation: (1) if recall is delayed by a distracting task, the recency effect disappears (the STM trace decays) but the primacy effect remains; (2) if encoding is slowed (giving more time per word), primacy improves but recency is unchanged.
The multi-store model treats short-term memory as a single, unitary store. Which patient study most directly challenged this assumption?
A: Patient H.M., who could not form new long-term memories but retained normal STM
B: Patient K.F. (Shallice & Warrington, 1970), who had severely impaired STM but near-normal long-term learning
C: Patient Clive Wearing, who had intact procedural memory but impaired episodic recall
D: Patient N.A., whose STM was intact but LTM for verbal material was selectively impaired
Patient K.F. sustained a motorcycle accident that severely damaged his phonological STM — his digit span was just one or two items. But crucially, his ability to learn new information into long-term memory was relatively intact. This was the reverse of H.M.'s pattern (intact STM, impaired LTM). If STM were a necessary gateway to LTM (as the model implies), K.F. should have been unable to form new long-term memories. The fact that he could suggested that STM and LTM are not linked in a strict serial gateway, and that STM itself is not unitary — prefiguring Baddeley's working memory model.
Flerlagringsmodellen
Atkinson and Shiffrin (1968) proposed that memory consists of three distinct stores. Which sequence correctly describes the flow of information through the multi-store model?
Om det här quizet
Atkinson and Shiffrin's multi-store model (1968) was the first systematic account of how information flows through distinct memory stores — from the fleeting sensory register through short-term memory to long-term storage. It generated decades of productive research into coding, capacity, and duration.
This quiz covers the three stores, the evidence that shaped them, and the model's limitations. Key sources include Sperling (1960) on iconic memory, Miller (1956) on STM capacity, Peterson and Peterson (1959) on STM duration, and Conrad (1964) on acoustic coding.