Memory- questions:
Short and long term memory:
1. Research into the capacity of the STM was investigated by Jacobs. He gave participants a string of
digits (numbers and letters) that they had to recall out loud. Each time they recalled it correctly, Jacobs
would add another digit onto the string. He repeated this until the participant failed to recall the list in
the correct order. He found that recall for numbers was 9.3 but for letters it was 7.3, perhaps because
there are more letters to learn than numbers. This shows that the capacity of the STM is around 7 +/- 2
chunks, as demonstrated by Miller’s research. He proposed chunking, which is that we can break
down information into meaningful, shorter chunks to remember it more easily.
One limitation of research into capacity is that the capacity of STM may be even more limited than initially
proposed. Cowan conducted a meta-analysis of many studies investigating capacity of the STM and found that
the average capacity is around 4 chunks. He also found similar results to the capacity for visual information
too, rather than just verbal stimuli. This is a weakness as it suggests that the capacity of STM may be
overestimated and not as extensive as first thought, and so Jacobs’ research must be used with caution.
Another limitation is that there may be individual differences, influencing how much information we retain in
STM. Jacobs found that the recall span gradually increased with age. People who were 8 years old
remembered an average of 6 chunks, whilst 19 years old remembered an average of 8.6 chunks. This shows
that biological factors such as brain capacity or the development of chunking might also influence capacity of
the STM. Therefore, this is a weakness as it shows that we need to take a more idiographic approach when
looking at the capacity of memory stores as it may differ with age, a confounding variable.
2. Research into the duration of the STM and LTM has been investigated by Peterson and Peterson and
Bahrick. For STM, Peterson and Peterson gave participants a trigram/consonant syllable to
remember (a three digit letter sequence). They had to recall it after a retention interval, in which they
had to count a three digit number sequence backwards for certain amounts of time, ranging from 3 to
18 seconds. They found that after 3 seconds, 90% successfully recalled the trigram, but after 18
seconds, only 2% were correct. This shows that the duration of the STM is extremely limited, only
around 18 seconds until it decays, if verbal rehearsal is prevented.
Research into the duration of the LTM was conducted by Bahrick, who gathered 392 participants. They had to
take part in a photo recognition test, where they would recall the names of 50 people in their yearbook, and a
free recall test, where they would recall the names of the fellow graduates. Bahrick tested them after 15 and
48 years. For the photo recognition task, he found that after 15 years, 90% remembered their graduates, but
after 48 years, that decreased to 70%. For the free recall task, after 15 years, around 60% recalled names
correctly, but that decreased to 30% after 48 years. Therefore, he found that the duration of the LTM is
potentially unlimited.
A limitation of research into STM is that the duration may be limited due to displacement and not decay,
which is not what Peterson and Peterson intended to study. When they were verbally recalling the numbers in
the retention interval, this information was being inputted to their STM, hence pushing the syllables out,
because of its limited capacity. Reitman confirmed this when instead of numbers, auditory tones were used,
which does not interfere with verbal rehearsal. He found that the duration of the STM was longer. Therefore,
Peterson and Peterson’s study demonstrated that the duration of the STM is limited because of displacement,
and so did not measure duration of the LTM effectively.
, Yet, a strength of Bahrick’s information on LTM is that it has high external validity, compared to many of the
other memory tasks, which are often contrived and artificial. His task, which involved recalling people from a
class yearbook can be generalised as it was meaningful and a task which would be seen in everyday life. His
experiment was also a field study, adding to the fact it has high external validity. However, a confounding
variable, such as looking at the yearbook or keeping in touch with friends could have weakened the validity of
his experiment, which is a limitation of his experiment, despite the fact that his findings can be representative
of the wider population.
3. Baddeley investigated coding into both short and long term memory. Baddeley used an independent
measures design, and had four different conditions, with either a semantically similar or dissimilar list of
words, and either a list of acoustically similar words. Participants in each group had to try and recall the
word lists either immediately, testing STM, or after a 20 minute interval, testing LTM. He found that
participants struggled the most recalling acoustically similar words immediately and semantically similar
words after the 20 minute interval. This shows that information is encoded acoustically for STM and
semantically for LTM.
One limitation of Baddeley’s research is that STM could actually be encoded visually. Brandimote found that
participants used visual coding in the STM if given a visual task to remember, when verbal rehearsal was
prevented (lalala). Usually, visual images would be ‘translated’ verbally but because it was prevented,
participants had to use visual codes. This shows that STM does not have to be exclusively encoded
acoustically but could be visually encoded too.
Another limitation of Baddeley’s research is that he used artificial, contrived stimuli, therefore his study has
low ecological validity, and his findings cannot be generalised beyond the laboratory. His participants had to
remember meaningless words which have no association with everyday life, hence cannot be generalised.
However, when we process meaningful information, such as people’s addresses, we may use semantic
coding in the STM. This is a limitation of Baddeley’s research as it suggests his findings have limited
applications, and there can be alternative ways of encoding STM.
Multi Store Model of Memory:
1. The multi-store model of memory was proposed by Attkinson and Shiffrin, who believed that memory
was divided into three separate stores, and information was passed through in a linear way. First,
environmental stimuli from the five senses is processed in the sensory register store, which has an
extremely short duration of around 500 milliseconds. Information is only transferred to the STM if it is
payed attention to. In the STM, information is processed, encoded acoustically and is transferred to
the LTM by maintenance rehearsal, which is repetition of information. Attkinson and Shiffrin proposed
there is a direct correlation between rehearsal in the STM and strength of the information retained in
the LTM. If it is to be transferred to the LTM, information is elaboratively rehearsed, which is where
new information processed is associated with previous information already in the LTM. If we want to
access information from the LTM, we have to pass it back through the STM- this is called retrieval.
2. One criticism of the multi-store model of memory is that STM may not be entirely separate. The MSM
claims that STM comes before LTM, however research suggests that this is not true. Logie proposed
that the STM actually relies on LTM, and that in order to chunk digits and numbers to remember them,
such as ITVBBCAAA, we need to associate this new information with information already stored in our
LTM, suggesting that STM and LTM are linked. This notion is further demonstrated by Ruchkin, who
gave participants a list of words and pseudo-words to remember. They found that participants had more
