Memory
, LT1
Features of Memory Stores
STM: Limited capacity (5-9 items), duration is 18-30 seconds, mainly coded acoustically
LTM: Duration of a lifetime, mainly coded semantically, capacity is unlimited
CODING CAPACITY DURATION
The form which information is stored in various ways: The amount of information that can be stored/held in a The length of time that information can be held in our
- Semantically memory store memory stores
- Acoustically Jacobs (1887) Digit Span (STM) Peterson and Peterson (1959)
- Visually He gave a list of digits they had to recall e.g.4 and if they Tested 24 undergraduates who each took part in 8
Baddeley (1966) got it right they would go to 5 until wrong trials. Given a trigram (3 letters) and asked to count
Gave different word lists to 4 groups (shown via a backwards in 3s from a 3 digit number e.g. 300
projector, each word for 3 secs) and asked them to Mean for digits: 9.3 (interference task- prevents rehearsal of the trigram).
remember them and recall them in order; immediately Mean for letters: 7.3 On each trial they were asked to stop at a different time
(STM) or after 20 minutes (LTM) - 3,6,9,12,15,18 (called the retention interval)
Miller (1956) Span of memory and chunking
Group 1: acoustically similar Found things come in 7s After 3 seconds= 80% trigrams recalled correctly
Group 2: acoustically dissimilar Avg capacity = 7 (7+/-2 = 5-9 items) 6 seconds= 50%
Group 3: semantically similar 18 seconds= less than 10%
Group 4: semantically dissimilar A03: Miller notes we can increase capacity by using Therefore the STM has a duration of 18-30 seconds
chunking by recalling 5 words/letters at a time to make
STM: worse on acoustic words it easier to recall e.g. phone numbers Bahrick et al (1975) - high school yearbooks test
LTM: worse on semantic words Studied 392 pps from Ohio aged 17-74
Shows that LTM codes semantically and STM codes JACOBS STUDY LACKED CONTROL OVER EXTRANEOUS 1. Photo recognition (50)
acoustically AND CONFOUNDING VARIABLES: Pps may have got Within 15 years of graduation: 90% accurate
distracted or may have been tired and therefore not After 48 years: 70%
HIGH CONTROL: same standardised procedures e.g. 3 have performed well = decreases replicability = 2. Free recall (name of graduating class)
seconds from projector = replicable and can be checked decreases credibility 15 years: 60%
for consistent results After 48 years: 30%
MILLER OVERESTIMATED CAPACITY OF STM: Cowan Therefore LTM has a long duration
LACKED MUNDANE REALISM: no relevance to real (2000) reviewed the research and said that capacity was ORDER EFFECTS: repeated measures design e.g. fatigue
memory tasks as it is artificial and not an everyday task only 4 chunks of information therefore Miller’s 5 was = decreases accuracy and confusion between tasks
= cannot be generalised and applicability can be more realistic than 7 and questions 7+/-2 HIGH EXTERNAL VALIDITY: meaningful task =
questioned increases generalisability as it doesn’t lack mundane
realism. However, confounding variables not controlled
as they could have recently looked at the book =
decreases credibility
, LT1
Features of Memory Stores
STM: Limited capacity (5-9 items), duration is 18-30 seconds, mainly coded acoustically
LTM: Duration of a lifetime, mainly coded semantically, capacity is unlimited
CODING CAPACITY DURATION
The form which information is stored in various ways: The amount of information that can be stored/held in a The length of time that information can be held in our
- Semantically memory store memory stores
- Acoustically Jacobs (1887) Digit Span (STM) Peterson and Peterson (1959)
- Visually He gave a list of digits they had to recall e.g.4 and if they Tested 24 undergraduates who each took part in 8
Baddeley (1966) got it right they would go to 5 until wrong trials. Given a trigram (3 letters) and asked to count
Gave different word lists to 4 groups (shown via a backwards in 3s from a 3 digit number e.g. 300
projector, each word for 3 secs) and asked them to Mean for digits: 9.3 (interference task- prevents rehearsal of the trigram).
remember them and recall them in order; immediately Mean for letters: 7.3 On each trial they were asked to stop at a different time
(STM) or after 20 minutes (LTM) - 3,6,9,12,15,18 (called the retention interval)
Miller (1956) Span of memory and chunking
Group 1: acoustically similar Found things come in 7s After 3 seconds= 80% trigrams recalled correctly
Group 2: acoustically dissimilar Avg capacity = 7 (7+/-2 = 5-9 items) 6 seconds= 50%
Group 3: semantically similar 18 seconds= less than 10%
Group 4: semantically dissimilar A03: Miller notes we can increase capacity by using Therefore the STM has a duration of 18-30 seconds
chunking by recalling 5 words/letters at a time to make
STM: worse on acoustic words it easier to recall e.g. phone numbers Bahrick et al (1975) - high school yearbooks test
LTM: worse on semantic words Studied 392 pps from Ohio aged 17-74
Shows that LTM codes semantically and STM codes JACOBS STUDY LACKED CONTROL OVER EXTRANEOUS 1. Photo recognition (50)
acoustically AND CONFOUNDING VARIABLES: Pps may have got Within 15 years of graduation: 90% accurate
distracted or may have been tired and therefore not After 48 years: 70%
HIGH CONTROL: same standardised procedures e.g. 3 have performed well = decreases replicability = 2. Free recall (name of graduating class)
seconds from projector = replicable and can be checked decreases credibility 15 years: 60%
for consistent results After 48 years: 30%
MILLER OVERESTIMATED CAPACITY OF STM: Cowan Therefore LTM has a long duration
LACKED MUNDANE REALISM: no relevance to real (2000) reviewed the research and said that capacity was ORDER EFFECTS: repeated measures design e.g. fatigue
memory tasks as it is artificial and not an everyday task only 4 chunks of information therefore Miller’s 5 was = decreases accuracy and confusion between tasks
= cannot be generalised and applicability can be more realistic than 7 and questions 7+/-2 HIGH EXTERNAL VALIDITY: meaningful task =
questioned increases generalisability as it doesn’t lack mundane
realism. However, confounding variables not controlled
as they could have recently looked at the book =
decreases credibility
