types of memory

• Capacity is how much can be stored
• Duration is how long things can be stored
• Coding is the format things can be stored in
• Sensory coding is when we store sensory information, this includes the acoustic code, which is what we use to store information about sound and the visual code which is what we use to store information about images
• Semantic coding is when we store information according to its meaning.
• Baddeley in 1966 conducted a laboratory experiment to investigate the coding in LTM and STM. He asked participants to recall four lists of words that either sounded similar, sounded different, had a similar meaning or a different meaning. He found participants recalling acoustically similar words had more difficulty when tested right after learning. 20 minutes after learning, participants recalling semantically similar words had more difficulty. He concluded acoustic coding is used in STM and semantic coding in LTM.
• The first memory store is the sensory register which holds sensory information while we process it, it has a large capacity and a short duration of 1 - 2 seconds. It uses the sensory code.
• We use the short term memory store to store information for a short period of time while we need it to complete an ongoing task. The capacity and duration of this store is limited. The coding can be either sensory or semantic but is mainly acoustic.
• The long term memory store has a very large capacity and a very long duration. It can hold information for many years and stores information in a semantic code.
• Atkinson and Shiffrin combines the sensory register, STM and LTM and came up with the multi store model. A key part was the transfer of information across the three memory stores. According to them, information is unidirectional and flows from the sensory register to STM and from STM to LTM.
• For information to be transferred from the sensory register to the short term memory we have to pay attention. For information to transfer from the short term memory store to the long term memory store we have to rehearse it.
• The multi store model predicts that each memory store can be damaged independently. If someone damages their LTM store they can still hold information in the sensory register and the STM
• According to the multi store model, If someone damages their STM they cannot form new long term memories.
• Sperling in 1960, had supporting evidence of the sensory register using a lab experiment. He presented pps with a grid of letters for 50 milliseconds then immediately tested their recall of the letters. In the first condition, pps were asked to recall as many letters as possible whilst the second condition had to recall a row in the grid but were not specified which row until after they had seen the grid.
• Sperling found the first group could recall an average of 5 letters and the second group could recall 3 letters. The second group had to store all 3 rows when asked to recall just one row. This suggests the sensory register has a large capacity but its duration is very short.
• Jacobs in 1887 has supporting evidence for capacity in the short term memory. He asked pps to recall a string of letters or numbers of different lengths. He found they could recall an average of 7 letters or numbers. He concluded STM has a limited capacity.
• George Miller did further research on STM, in the 1950s and found that the maximum capacity for STM was 7 + / - 2 chunks of information. This supports Jacobs findings about the capacity of STM being limited but we can hold more if we combine information into meaningful chunks.
• Evidence for duration comes from Bahrick in 1975. He studied whether pps could remember the names of their highschool classmates and tested both recall and recognition. He found that 15 years after pps left school that their memory recall was 60% accurate and their recognition was 90% accurate. 48 years after school pps memory recall dropped to 30% but their recognition was 80%. So he found that LTM has a very long duration but our ability to retrieve information is better using recognition
• Bahrick's study had high ecological validity as information came from pps real life. However he could not control extraneous variables.
• Henry Molaison (HM) was a case study that supports the multi store model. He had damage to his LTM but still had a functional STM. This shows we can damage LTM without damaging retrieval ability in the STM. This supports the idea we have separate stores for our STM and our LTM.
• Brain imaging studies show different parts of the brain are active when we hold information for short or long periods of time. The frontal cortex is active when we retain information for a short period of time and the hippocampus is active when we retain information for a long period of time.