Short Term Memory (STM) Long Term Memory (LTM)
Capacity "Magic number 7 plus or Assumed unlimited
(how much data is stored) minus 2" - 5-9 items
Duration 18-20 seconds Possibly a lifetime
(how long data is stored)
Encoding Acoustic Semantic
(the form in which data is
STM Capacity - Miller (1956): Concluded that the capacity of STM was five to nine items, but showed how chunking could increase this by grouping information into meaningful units using established memory stores.
STM Duration - Peterson and Peterson (1959): Read out nonsense trigrams to Ps, (eg. XJP), then delayed recall by asking them to counting backwards from between 3 and 18 seconds. 90% of trigrams were recalled after 3 seconds, but only 5% after 18 seconds which suggests the duration of STM is between 18 and 20 seconds.
STM Ecoding - Baddeley (1966): Ps were divided into four groups, each group heard a list of words from one of the following categories; acoustically similar, acoustically dissimilar, semantically similar and semantically dissimilar words. After hearing the words they were asked to recall in the right order. Similarity of words had very little detrimental effect. Acoustically similar words were much harder to recall in the right order (55% accurate) then acoustically dissmilar words (75% accurate). Suggesting STM relies more on the sound of the words than their meaning.
LTM Duration - Bahrick et al (1975): Graduates were split into two groups shown photographs from their high-school yearbooks. One group had a list of names and had to match pictures to the names, the other group had no name cues. The group with cues had 60% accurate recall after 47 years. The group without cues was less than 20% accurate after the same amount of time. This shows that people can rmember certain types of information for almost a lifetime.
LTM Encoding - Baddeley (1966): Similar process to STM Encoding. Except Ps were given an interval of 20 minutes, during which they had a distraction task to carry out, before asked to recall the words in the correct order. Recall was much worse for semantically similar words (55% accuracy) than for semantically dissimilar words (85% accuracy) suggesting that LTM…