Coding, Capacity and Duration
Baddeley (1966) - CODING
He gave a list of words to 4 groups of participants.
The lists were either acoustically similar, semantically similar, acoustically dissimilar or semantically dissimilar.
They had to recall the words in the correct order either immediately or after a 20 minute task.
IMMEDIATE RECALL - PPs were worse recalling acoustically similar words. This suggests that information is coded acoustically in
STM
RECALL AFTER 20 MINS - PPs were worse recalling semantically similar words. Suggests info is coded semantically in LTM.
EVALUATION: Lacks ecological validity
There are other types of LTM and other types of coding this study didn't consider
The study used an independent groups design, so there was no control of participant variables
Sperling 1960 - Sensory Register
In a laboratory experiment, PPs were shown a grid with 3 rows of 4 letters for 0.05 seconds. They then had to immediately recall
the whole grid or a randomly chosen grid
When PPs had to recall the whole grid, they recalled ⅘ letters. When a particular row was selected, PPs recalled an average of 3
items, no matter which row was selected
The PPs didn't know which row would be selected, so it could be concluded that they would have been able to recall 3 items from
any row, therefore almost the whole grid was in their sensory register
They couldn't report the whole grid because the trace faded before they could finish recall
Jacobs 1887 - Capacity of STM (digit span) Miller 1956 – Capacity of STM (chunking)
PP given 4 digits and asked to recall them in correct order out loud He noticed that people could remember about 7 items,
numbers with more ease than letters
If they recall this correctly, the number of digits increases one by
one until they can't correctly recall order He suggested that the capacity of STM is 7 ± 2 – this
number is called Millers magic number
The mean digit (number) span was 9.3 items, for letters it was 7.3
He suggested that we use chunking to combine individual
The capacity increased with age during childhood
letters or numbers into larger and more meaningful units,
This concluded that STM has a limited storage capacity of 5-9 items. as people can recall 5 words as well as they can 5 letters
It also showed that memory techniques such as chunking help recall The STM can hold about 7 pieces of chunked information,
which increases the capacity of STM.
Digits may have been easier to recall as there's only 10 different
digits compared to 26 letters.
EVALUATION: Artificial and lacks ecological validity - not real life. If Cowan 2001
the information was meaningful, it may have been recalled better,
Cowan (2001) reviewed Millers research. He claimed that
so STM could have an even bigger capacity
STM can actually only hold about 4 pieces of chunked
Confusion - The previous sequences the PPs recalled could have information
confused them on future trials
This means that the lower end of Miller’s magic number is
, more appropriate according to Cowan.
Peterson and Peterson 1959 – Duration of STM (trigrams)
They tested 24 undergrad students. They took part in 8 trials. In each trial, students were given a constant syllable (trigram) such as
YGC to remember. They were also given a 3 digit number to count backwards in 3’s from (an interference task to prevent rehearsal
of trigram). On each trial they were asked to stop after certain amount of time (called the retention interval)
After 3 seconds, PPs could only recall 80% of the trigrams correctly. After 18 seconds, only 10%. This suggests that when rehearsal
is prevented//[=, very little stays in the STM for longer than 18 seconds.
EVALUATION: It was a laboratory experiment (reliable, good control of variables); it lacks ecological validity (nonsense trigrams are
artificial, meaningful memories may last longer in STM); Only 1 type of stimulus is used (different stimuli may last for different
period of time in STM); Each person saw many trigrams, which could have caused confusion, meaning that only the first trigram
recalled was a realistic trial
Bahrick et al 1975 – Duration of LTM (Yearbook photos)
Studied 392 participants from Ohio, aged between 17 & 74. They obtained their high school yearbooks.
They tested the PPs recall in various ways:
1. Photo Recognition Test – Consisted of 50 photos, some of them from the PPs yearbook, and they had to recall their
names
2. Free Recall Test – Asked to recall names of their graduating class
Those tested 15yrs after leaving school were 90% accurate on photo recognition test, 60% on free recall
After 48ys, recall declined to 70% for photo recognition, 30% for free recall
This is evidence of VLTMs in real life settings. Recognition is better than recall, which could suggest that we have a huge store of
information that isn’t always easily accessible – we need help to remember it.
EVALUATION: It was a field experiment so has high ecological validity. However not all variables were controlled
It showed better recall than other studies on LTM, suggesting meaningful info is stored better
This type of info could have been rehearsed (EG if still in touch with classmates) which would increase recall. So results can’t be
generalized.
Multi store Model - Atkinson & Shiffrin (1968)
Craik & Watkins 1973 - More than one type of rehearsal Shallice & Warrington 1974 - KF Case study
Contrary to the MSM, Craik and Watkins found that how likely a Shallice and Warrington studied a patient with amnesia (KF)
memory is to be transferred to the LTM depends on the type of and found him to have poor STM for digits that were read to
rehearsing, rather than the amount of rehearsal. him, but he could recall better if he read the digits.
They discovered 2 types of rehearsal: This could suggest that there are separate stores for different
Baddeley (1966) - CODING
He gave a list of words to 4 groups of participants.
The lists were either acoustically similar, semantically similar, acoustically dissimilar or semantically dissimilar.
They had to recall the words in the correct order either immediately or after a 20 minute task.
IMMEDIATE RECALL - PPs were worse recalling acoustically similar words. This suggests that information is coded acoustically in
STM
RECALL AFTER 20 MINS - PPs were worse recalling semantically similar words. Suggests info is coded semantically in LTM.
EVALUATION: Lacks ecological validity
There are other types of LTM and other types of coding this study didn't consider
The study used an independent groups design, so there was no control of participant variables
Sperling 1960 - Sensory Register
In a laboratory experiment, PPs were shown a grid with 3 rows of 4 letters for 0.05 seconds. They then had to immediately recall
the whole grid or a randomly chosen grid
When PPs had to recall the whole grid, they recalled ⅘ letters. When a particular row was selected, PPs recalled an average of 3
items, no matter which row was selected
The PPs didn't know which row would be selected, so it could be concluded that they would have been able to recall 3 items from
any row, therefore almost the whole grid was in their sensory register
They couldn't report the whole grid because the trace faded before they could finish recall
Jacobs 1887 - Capacity of STM (digit span) Miller 1956 – Capacity of STM (chunking)
PP given 4 digits and asked to recall them in correct order out loud He noticed that people could remember about 7 items,
numbers with more ease than letters
If they recall this correctly, the number of digits increases one by
one until they can't correctly recall order He suggested that the capacity of STM is 7 ± 2 – this
number is called Millers magic number
The mean digit (number) span was 9.3 items, for letters it was 7.3
He suggested that we use chunking to combine individual
The capacity increased with age during childhood
letters or numbers into larger and more meaningful units,
This concluded that STM has a limited storage capacity of 5-9 items. as people can recall 5 words as well as they can 5 letters
It also showed that memory techniques such as chunking help recall The STM can hold about 7 pieces of chunked information,
which increases the capacity of STM.
Digits may have been easier to recall as there's only 10 different
digits compared to 26 letters.
EVALUATION: Artificial and lacks ecological validity - not real life. If Cowan 2001
the information was meaningful, it may have been recalled better,
Cowan (2001) reviewed Millers research. He claimed that
so STM could have an even bigger capacity
STM can actually only hold about 4 pieces of chunked
Confusion - The previous sequences the PPs recalled could have information
confused them on future trials
This means that the lower end of Miller’s magic number is
, more appropriate according to Cowan.
Peterson and Peterson 1959 – Duration of STM (trigrams)
They tested 24 undergrad students. They took part in 8 trials. In each trial, students were given a constant syllable (trigram) such as
YGC to remember. They were also given a 3 digit number to count backwards in 3’s from (an interference task to prevent rehearsal
of trigram). On each trial they were asked to stop after certain amount of time (called the retention interval)
After 3 seconds, PPs could only recall 80% of the trigrams correctly. After 18 seconds, only 10%. This suggests that when rehearsal
is prevented//[=, very little stays in the STM for longer than 18 seconds.
EVALUATION: It was a laboratory experiment (reliable, good control of variables); it lacks ecological validity (nonsense trigrams are
artificial, meaningful memories may last longer in STM); Only 1 type of stimulus is used (different stimuli may last for different
period of time in STM); Each person saw many trigrams, which could have caused confusion, meaning that only the first trigram
recalled was a realistic trial
Bahrick et al 1975 – Duration of LTM (Yearbook photos)
Studied 392 participants from Ohio, aged between 17 & 74. They obtained their high school yearbooks.
They tested the PPs recall in various ways:
1. Photo Recognition Test – Consisted of 50 photos, some of them from the PPs yearbook, and they had to recall their
names
2. Free Recall Test – Asked to recall names of their graduating class
Those tested 15yrs after leaving school were 90% accurate on photo recognition test, 60% on free recall
After 48ys, recall declined to 70% for photo recognition, 30% for free recall
This is evidence of VLTMs in real life settings. Recognition is better than recall, which could suggest that we have a huge store of
information that isn’t always easily accessible – we need help to remember it.
EVALUATION: It was a field experiment so has high ecological validity. However not all variables were controlled
It showed better recall than other studies on LTM, suggesting meaningful info is stored better
This type of info could have been rehearsed (EG if still in touch with classmates) which would increase recall. So results can’t be
generalized.
Multi store Model - Atkinson & Shiffrin (1968)
Craik & Watkins 1973 - More than one type of rehearsal Shallice & Warrington 1974 - KF Case study
Contrary to the MSM, Craik and Watkins found that how likely a Shallice and Warrington studied a patient with amnesia (KF)
memory is to be transferred to the LTM depends on the type of and found him to have poor STM for digits that were read to
rehearsing, rather than the amount of rehearsal. him, but he could recall better if he read the digits.
They discovered 2 types of rehearsal: This could suggest that there are separate stores for different
