STM and LTM: capacity, duration and coding

what is capacity?

the amount of information the memory store holds

capacity of the STM

research suggests STM capacity is limited to 7+ / -2 items (or between 5 and 9 units). when new information comes into STM it pushes out the old information (displaces) due to this limited capacity

research into STM capacity

miller (1956) supports the idea of a limited capacity with his research using the digit span technique. this involves reading series of digit sets that get progressively longer. the individual is asked to immediately repeat the digit set back in the right order. miller found that participants could recall on average 7+ / -2 digits. he concluded that STM has a limited capacity and that new information coming into STM ‘pushes out’ (displaces) the old information due to this limited capacity

capacity of the LTM

research suggests that the capacity of LTM is unlimited. as capacity of the LTM is notoriously difficult to measure, there is very little research in this area.

what is duration?

the length of time the memory store holds information

duration of the STM

research (such as that conducted by peterson and peterson) has suggested that information can only be held in STM for a limited period of time (18-30 seconds). however, this can be extended through rehearsal (i.e. repeating information over and over again)

research into STM duration

peterson and peterson (1959) asked 24 students to listen to a ‘nonsense trigram’ (a random string of 3 letter such as ‘WRT’). immediately after hearing the trigram, they heard a random 3-digit number and they were asked to count backwards in 3s from this number in order to prevent rehearsal of the trigram. they were asked to recall the trigram after 3, 6, 9, 12, 15 and 18 seconds. it was found that the highest level of recall was after 3 seconds (90% recalled the trigram) but this was decreased rapidly as the duration increased until there was only 2% recalled after 18 seconds

duration of the LTM

research suggests that duration of LTM is potentially unlimited (can potentially last up to a lifetime). even if information cannot be retrieved that does not mean the information is not stored in LTM

research into LTM duration

bahrick et al (1975) asked 392 participants ages 17-74 to name old classmates from their high school (free recall test). they were giving 50 photos, some from the participants’ high school yearbook and asked if they could recognise their classmates from these photos (recognition test). they found that 15 years after graduation, free recall was about 60% accurate. after 48 years this dropped to 30%. recognition was 90% accurate after 15 years, dropping to about 70% after 48 years. this suggests LTM could have a potentially unlimited duration (LTM could potentially last up to a lifetime) and that even if a long-term memory cannot be freely recalled, this does not mean it is no longer stored, it may just need assistance to be retrieved (i.e. recognition rather than free recall)

what is coding?

the way information is processed so it can be stored in memory

coding of the STM

research suggests that encoding in STM is mainly acoustic ( i.e. by sound). when a person is presented with a list of items to remember, they will try to hold them in STM by saying them over and over out loud

coding of the LTM

research suggests encoding in LTM is mainly semantic (by meaning). we need to understand something (process its meaning) for it to be retained in LTM

research into STM and LTM coding: baddeley (1966)

procedure: 72 participants randomly allocated to one of four conditions (independent group design). each group was shown a list of ten words, the words were either acoustically similar (e.g. cat, cab, can), acoustically dissimilar (e.g. good, huge, hot). all words were one-syllable (as control)

STM coding: recall was tested immediately. the number of words recalled in the correct order was measured

LTM coding: recall was tested following a delay/20 minutes

STM findings: recall of acoustically similar words was poorer when tested immediately/in STM compared to acoustically dissimilar words

he concluded that this was because there was confusion based on the way the words sounded. this suggests that STM encodes information acoustically

LTM findings: recall of semantically similar words was poorer when tested after a delay of 20 minutes/in LTM compared to semantically dissimilar words

he concluded that this was because there was confusion based in the meaning of the words. this suggests that LTM encodes information semantically