Class Notes (1,100,000)
CA (630,000)
Western (60,000)
PSYCH (7,000)

Psychology 2135A/B Lecture Notes - Cognitive Psychology, Clive Wearing, Behaviorism

Course Code
Robert Brown

of 5
Lecture 7 - Short Term Memory
The case of Clive Wearing
- random convulsions, except in music; can still play piano (different kind of memory that has
been preserved)
- could still remember to play music and sing, but no longer able to remember things for over 20-
30 seconds
- very little memory before and after illness
- experiences things are like just waking up
- declarative memory (things able to talk about); procedural memory (things know how to do,
usually survives)
- recognizes wife but not children; like seeing her for the first time in a long time
- some manners have survived
- what life would be like if all we had was STM
- lost some semantic memory, even those prior to injury; cannot learn new semantic memory
- brain scans show some loss of fluid and other parts
William James on memory
- see quote on WebCT
Donald Hebb on memory
- 1949, the Cell Assembly model
- neurons that fire together wire together
- every time neuron A sends a signal to neuron B as part of response to a stimulus, connection
between two becomes stronger
- stronger connection means stronger signal from A arriving at B, so it is more likely that B will
fire as a result; more likely that both fire together
- two ways to represent experience:
- which neurons are active right now in response to what just happened
- state (what goes on right now)
- strengths of connections between neurons, which reflect experience (learning)
- trait (structural changes over a period of time due to experience)
- strengths of connections = LTM
- LTM is structural changes in the nervous system
- active right now = STM
- STM = changes in activation level
George Miller (1956) on short-term memory
- the magical number seven, plus or minus two
- capacity of STM
- consistent finding of a limit on human ability to remember information
- but amount of information in immediate memory can be increased through chunking
- chunk is a collection of elements strongly associated with one another but weakly associated
with elements in other chunks; use knowledge in world to make connections
- reduces the amount of information to be stored
Why did researchers come to believe in STM independent of LTM?
- Arthur Melton (1963) started formal arguments for the existence of STM at the APA annual
- theory of STM is distinct from LTM is currently controversial; some people reject theory
(Sederberg, Howard, Kohana, 2008)
Memory loss following brain damage
- can get new information into STM but not LTM?
- HM is most famous patient
- had access to some of old memories (childhood); could not learn new memories
- surgery to relieve severe epilepsy in 1953, at age 27
- bilateral excision of medial temporal lobe
- "it's like waking from a dream"
- anterograde amnesia: inability to remember things that happened after brain damage. Implies
inability to transfer new memories from STM into LTM
- retrograde amnesia: inability to remember things that happened before brain damage. Implies
inability to retrieve existing memories from LTM
- Scoville and Milner (1957); after surgery, profound anterograde amnesia; also showed some
retrograde amnesia
- studies by Brenda Milner:
- good vocabulary and language, normal IQ, no attention disorder
- things unable to learn:
- where he lived, who cared for him, what he ate at his last meal, the year, president, age,
failed to recognize himself on his 40th birthday in 1982
- somehow learned that, in the 1970s sitcom All in the Family, Archie Bunker's son-in-law was
called "Meathead"; possibly done by residual of hippocampus
- could not learn declarative tasks: photographs of people, verbal material, sequences of digits,
nonsense patterns, complex geometric designs, could not expand his digit span
- showed procedural learning: improved with practice (procedural memory) but denied having
that practice (declarative memory)
- Milner (1962) trained HM on a mirror-drawing task
- Cohen and Corkin (1981); HM showed procedural learning on Tower of Hanoi puzzle
- consensus was that LTM was impaired but his STM spared
- capacity of LTM is essentially infinite (can always learn), what about STM?
- Shepard and Teghtsoonian (1961); present 200 three-digit numbers in a row
- task: say when you hear a repeated number
- IV: interval before repetition; DV: probability of noticing repetition
- forgetting function: how does probability of noticing repetition vary with interval?
- are there separate forgetting functions for STM and LTM
- probability of noticing repetition fell dramatically at first
- steep decline in probability ended at interval = 7 items (between first and second
- probability of noticing then fell more gradually, asymptoting at 60%
- initial steep decline occurs because response comes from STM (decays quickly)
- more gradual decline occurs when response depends upon LTM (slow decay, if at all)
- different slopes represent different forgetting functions so different storages
Why should STM have so small a capacity?
- sensory and LTM both have large capacity
- if STM was any larger, it would take too long to search through; need it fast
- LTM lasts a long time (Bahrick 1984 - permastore)
- Brown (1958) and Peterson and Peterson (1959)
- task: subjects see a stimulus and have to recall it after an interval
- rehearsal is prevented by having them count backwards during retention interval
- IV: length of interval in seconds
- result: interval longer than 18 seconds, no longer remember
- memory system where things must be rehearsed or are soon lost, but do not rehearse in
LTM so STM must be different from LTM
- criticism: counting backwards may interfere with STM
- STM is at most 30 seconds
Type of Code
- every stimulus has multiple aspects; colour, brightness, shape, surface texture, category, name;
All found in LTM, are all in STM?
- Brown/Peterson paradigm
- task: subjects briefly see a stimulus and have to recall it after an interval
- most of the errors made were phonological i.e. say P instead of T
- errors based on shape were rare i.e. say C instead of O
- no semantic errors observed (not possible anyways)
- STM uses only phonological code
Serial Position Effect
- in ordered recall task, subjects try to recall list of words in the order they were given
- accuracy higher for beginning (primacy effect) and end (recency effect) of list, lower for
middle of list
- primacy produced by LTM, recency reflects items still in STM
Mechanism of Loss
- how are things lost? Decay, interference, retrieval failure (retrieval cue is not good enough)?
- originally, LTM loss was blamed on interference and STM loss was blamed on decay
- interference: disruption of memory trace by other traces; degree of interference depends upon
similarity of the two memory traces
- decay: automatic fading of memory trace due to time
What do we think of those reasons now?
Type of Code
- we now know that STM can contain any kind of code (Brooks 1968, and Wickens 'Release
from Proactive Inhibition')
- Proactive inhibition: Previously learned information interferes with retrieval of more
recently learned information