Feb-22
2. Memory
Coding, capacity, and duration:
Coding: the format of information stored in a memory store
STM = mainly acoustic
LTM = mainly semantic
Baddeley (1966): similar/dissimilar words
1. Gave 4 groups of ppts lists of words. The lists were either acoustically/semantically similar,
or acoustically/semantically dissimilar.
2. He asked ppts to recall their word list in order immediately after hearing it, as well as 20
minutes after hearing it.
When recalling immediately, the acoustically similar group did the worst, suggesting that STM
codes acoustically (found it hard to distinguish between the different words because they
sounded similar, so made mistakes).
When recalling after 20 minutes, the semantically similar group did the worst, suggesting that
LTM codes semantically (found it hard to distinguish between the different words because they
had similar meanings, so made mistakes).
Capacity: how much information can be stored in a memory store.
STM = 5-9 items
LTM = unlimited
Jacobs (1887): digit-span
1. Gave 4 digits for ppts to recall aloud in order.
2. Increased number of digits to 5, 6, 7 items until ppts could no longer recall them correctly.
Mean span of digits recalled by ppts was 9.3.
Mean span of letters recalled by ppts was 7.3.
Miller (1956): chunking
o Observed that things often come in 7s in everyday life (e.g., music notes, days etc).
Concluded that capacity of STM is about 7 (+/-2) items.
o Also noted that people could recall 5 words equally as well as they can 5 letters by chunking
(i.e., grouping of digits/letters into units).
Duration: how long information can be stored in a memory store
STM = 18 seconds without rehearsal
Peterson (1959): consonant recall
1. Gave ppts consonant syllable (e.g., YCG) and a 3-digit number to count backwards from (to
prevent maintenance rehearsal).
2. 8 trials; stopped after a certain amount of time (3, 6, 9 … 18 seconds).
Found that STM has a short duration (without rehearsal) of 18 seconds.
, Feb-22
LTM = many years
Bahrick et al (1975): yearbook recall
1. 394 high school graduates (17-74-year-olds). Asked them to recognise 50 photos and recall
names of their graduate class.
Those tested within 15 years of graduation were 90% accurate for photo recall and 60% for free
recall.
Those tested after 48 years were 70% accurate for photo recall but only 30% for free recall.
Therefore, concluded that LTM lasts a long time. Very large duration.
EVALUATION:
Capacity:
- Digit span may have flawed methodology
o Very old study (19th century); likely lacked control due to confounding variables.
o Therefore, low validity.
HOWEVER: + Later research supports findings. Therefore, increased reliability and validity.
- STM span may be overestimated.
o Cowan (2001): reviewed research, concluding that STM was only 4 chunks, not 5-7.
o Therefore, lower end of Miller’s research may be more accurate.
Coding:
- Artificial stimuli used in research.
o Baddeley used word lists; we may code differently in real life.
o Therefore, low generalisability and ecological validity.
Duration:
- Artificial stimuli
o Peterson – consonant recall
o Therefore, low generalisability and ecological validity.
+ Real-life stimuli
o Bahrick – yearbook recall; use of meaningful memories.
o Therefore, high generalisability to everyday life and high ecological validity.
Multi-store model (MSM):
maintenance
rehearsal
Sensory register
Iconic
Stimulus STM LTM
Echoic
Other sensory
remembering decay decay
decay
2. Memory
Coding, capacity, and duration:
Coding: the format of information stored in a memory store
STM = mainly acoustic
LTM = mainly semantic
Baddeley (1966): similar/dissimilar words
1. Gave 4 groups of ppts lists of words. The lists were either acoustically/semantically similar,
or acoustically/semantically dissimilar.
2. He asked ppts to recall their word list in order immediately after hearing it, as well as 20
minutes after hearing it.
When recalling immediately, the acoustically similar group did the worst, suggesting that STM
codes acoustically (found it hard to distinguish between the different words because they
sounded similar, so made mistakes).
When recalling after 20 minutes, the semantically similar group did the worst, suggesting that
LTM codes semantically (found it hard to distinguish between the different words because they
had similar meanings, so made mistakes).
Capacity: how much information can be stored in a memory store.
STM = 5-9 items
LTM = unlimited
Jacobs (1887): digit-span
1. Gave 4 digits for ppts to recall aloud in order.
2. Increased number of digits to 5, 6, 7 items until ppts could no longer recall them correctly.
Mean span of digits recalled by ppts was 9.3.
Mean span of letters recalled by ppts was 7.3.
Miller (1956): chunking
o Observed that things often come in 7s in everyday life (e.g., music notes, days etc).
Concluded that capacity of STM is about 7 (+/-2) items.
o Also noted that people could recall 5 words equally as well as they can 5 letters by chunking
(i.e., grouping of digits/letters into units).
Duration: how long information can be stored in a memory store
STM = 18 seconds without rehearsal
Peterson (1959): consonant recall
1. Gave ppts consonant syllable (e.g., YCG) and a 3-digit number to count backwards from (to
prevent maintenance rehearsal).
2. 8 trials; stopped after a certain amount of time (3, 6, 9 … 18 seconds).
Found that STM has a short duration (without rehearsal) of 18 seconds.
, Feb-22
LTM = many years
Bahrick et al (1975): yearbook recall
1. 394 high school graduates (17-74-year-olds). Asked them to recognise 50 photos and recall
names of their graduate class.
Those tested within 15 years of graduation were 90% accurate for photo recall and 60% for free
recall.
Those tested after 48 years were 70% accurate for photo recall but only 30% for free recall.
Therefore, concluded that LTM lasts a long time. Very large duration.
EVALUATION:
Capacity:
- Digit span may have flawed methodology
o Very old study (19th century); likely lacked control due to confounding variables.
o Therefore, low validity.
HOWEVER: + Later research supports findings. Therefore, increased reliability and validity.
- STM span may be overestimated.
o Cowan (2001): reviewed research, concluding that STM was only 4 chunks, not 5-7.
o Therefore, lower end of Miller’s research may be more accurate.
Coding:
- Artificial stimuli used in research.
o Baddeley used word lists; we may code differently in real life.
o Therefore, low generalisability and ecological validity.
Duration:
- Artificial stimuli
o Peterson – consonant recall
o Therefore, low generalisability and ecological validity.
+ Real-life stimuli
o Bahrick – yearbook recall; use of meaningful memories.
o Therefore, high generalisability to everyday life and high ecological validity.
Multi-store model (MSM):
maintenance
rehearsal
Sensory register
Iconic
Stimulus STM LTM
Echoic
Other sensory
remembering decay decay
decay
