Human memory

1
Human memory

• How do humans learn?
• We learn by processing information and storing it
  into memory.
• Research on human learning has been directed
  toward understanding human memory.

2
How do we process information?

• Proceed from transient to permanent memory
  stores.
• Transient memory stores.
• Sensory memory
• Short-term (working memory)

3
Sensory memory

• What is?
• A buffer which holds sensory information long
  enough to identify what we are perceiving.
• Sensory specific.
• Last only for a brief period of time.

4
EvidenceSperling (1960)

• Studied span of apprehension.
• How much information can we process with one
  brief glance?
• Only 4 to 5 letters out of 3 x 4 letter array.
• But, subjects claimed .
• They can do better.
• They see the array but not enough time to report
  all the letters.

5
New technique

• If not enough time, reduce the time it takes to
  report.
• Use partial report
• Array - Signal which row - report
• Even though the signal was given after the
  stimulus was turned off.
• Report - 70

6
How long does it last?

• Sperling
• Delayed the signal
• 0, 150 ms, 300ms, 1s
• Found that up to 300ms, partial gt whole
• but at 1 s, partial whole
• Last about 300 ms
• Neisser - iconic memory

7
Auditory sensory memory

• How to study?
• Three-eared man technique
• Presented letters from three directions
• left, center, right
• Compared partial and whole report
• partial - signaled to report from which side.

8

• Found - partial is better than whole
• up to about 2 s
• controversial
• depends on the complexity of the stimuli
• Neisser - echoic memory

9
Suffix effect

• Another technique
• What would happen when additional information
  enters into sensory memory?
• Crowder and Morton (1969)
• presented digits
• asked subjects to say 0 at the end
• impaired their recall of digits
• but if you delay 0 - no interference

10
Conclusion

• There seems to be a very brief memory store at
  the beginning of information processing.
• Some still doubt its existence and its importance.

11
Short term memory

• Many suspected that the next store is short term
  store.
• Number of characteristics
• limited capacity
• fast rate of forgetting
• use of acoustic code
• neuropsychological evidence
• Many now consider STS to be a too simplistic idea

12
Limited capacity

• Memory span (Miller, 1956)
• 4235
• 3513679
• 837263857
• 94837436176354
• Capacity is 7 plus minus 2
• but, you can increase this by chunking

13

• chunking - use pre-exiting knowledge
• Why pre-existing knowledge influence capacity?
• Maybe this is not a separate store.
• Limited capacity --- weak evidence for a
  separate store
• Maybe there is a limit in how much information
  you can make it conscious.

14
Fast Forgetting Rate

• Peterson Peterson (1959)
• asked subjects to remember a trigram (CXT)
• distractor task (0 to 20 seconds)
• recall the trigram
• found
• forgetting occurred immediately
• nothing remained after 18 seconds

15
But..

• This is average performance over many trials.
• If you look each trial
• on 1st trial - almost no forgetting
• as trial progress - getting worse and worse
• Not retention interval
• Proactive interference - the more you learn, the
  more difficult it becomes to learn new material

16

• Very similar to forgetting in long term memory.
• Why does it disappear so fast?
• Maybe we did not encode it well enough
• Levels of Processing model

17
Coding difference

• Atkinson and Shiffrin (1968)
• STS - acoustic
• LTS - semantic
• Experiment (Conrad, 1964)
• presented letters (E D B Q T S)
• immediate recall
• found
• worse performance when letters sounded similar.
• Acoustic confusion - G Z

18
But.

• Acoustic code isnt the only one
• Potter Lambardi (1990)
• The knight rounded the palace searching for a
  place to enter. --- castle
• Immediate recall showed semantic confusion
• The knight rounded the castle searching for a
  place to enter.

19
So...

• Semantic code is also used in STS
• Similar to LTS
• Use other codes too
• similar movements in ASL cause confusion for
  hearing impaired people.
• cherological code.
• Different codes are represented in STS

20
Neuropsychological evidence

• Milner (1960)
• Observed HM
• Undergone bilateral lesion of parietal lobe.
• Could remember things for a few seconds.
• Could not retain for a long term.
• Why?
• Could not transfer information from STS to LTS

21

• Shallice Wallington (1970)
• studied K.F.
• impaired STS - two items on digit span
• no impairment on LTS - shows the normal primacy
  effect on a serial position curve.
• If STS comes before LTS, why didnt KF have
  impaired long term memory?

22
Conclusion

• Many doubts that STS is a separate from LTS.
• Some considers STS is a part of more elaborate
  system called working memory.