Ebbinghaus began the systematic study of memory using nonsense syllables. He showed that memory declined very rapidly at first, then levelled off. James (1890) observed that whilst memory appears to store some information for a lifetime other information is lost really quickly. He distinguished between two types of memory primary structure and secondary structure. Today these are known as short-term memory (STM) and long term memory (LTM).
The limited capacity of STM can be increased through chunking as Miller (1956) discovered. However, chunking depends on matching incoming information and it’s representation in LTM as Miller and Selfridges discovered (1950). The closer a sentence is to English the better it was recalled in the study. Bower and Springston (1970) presented a group of American students with familiar letters (e.g., fbi, phd0 and they grouped them in a different combination (egf, bip, hd) and asked other students.
More letters were recalled correctly when they were grouped into acronyms. The first group could chunk the information together and get it out of their mental dictionaries. Coding in STM is mainly acoustic as indicated by acoustic errors when Conrad (1964) did a study. 62 people thought they heard a B but it was a V. Semantic and visual coding are also used. The Brown-Peterson technique shows that STM’s duration is very short in the absence of rehearsal and especially when something else distracts the participant they forget again. If you keep rehearsing information then it can be stored in the STM.
There is no evidence how big LTM is. Lots of psychologists believe that it has an unlimited capacity. It has much longer duration than STM with memories stored in a more permanent way. Coding in LTM is mainly semantic but information still goes in visually and acoustically. Smells and tastes are also stored in LTM making it a flexible system. Atkinson and Shiffrin’s (1968, 1971) multi-store model of memory sees sensory memory, STM and LTM as permanent structural components of the memory system.
Rehearsal is a key control process as it acts as a buffer between sensory memory and LTM and helps the transfer of information to LTM. The recency effect reflects recall from STM and LTM. Both together makes up the serial position effect. Studies have been carried out on amnesiacs. Alcoholics suffering from Korsakoff’s syndrome have an intact STM as they can have a conversation with you but there LTM is bad and they can’t remember when the conversation was. This suggests that STM and LTM are separate storage systems.
Baddeley and Hitch’s working memory model believed that STM was more complicated than just a ‘stopping off station’ for information. STM is seen as the central executive who controls activities such as decision making and problem solving. Capacity is limited but flexible and can process information in any sense modality. The articulatory loop is like a verbal rehearsal loop. Information is represented, as it would be spoken the loop has two components: an artculatory control process based on inner speech and a phonological store holding on to a speech based information. Children who have difficulty learning to read may show some deflect in their phonological loops. The visuo-spatial scratch pad is the visual information. IT deals with the visual features e.g. size, colour shape. It is known as the inner eye.
Craik and Watkins distinguished between maintenance rehearsal and elaborate rehearsal. Maintenance rehearsal is the rehearsal that appears in the multi-storey model which is that the material is rehearsed as it was when it came into the memory whereas elaborate rehearsal is when the material is related to previous knowledge that is stored in LTM. Craik and Lockhart’s level of processing model (LOP) states that memory is a by-product of perceptual analysis. When information is taken into the memory it analyses it on various levels in the central processor. 1. Shallow level e.g. whether the word is small or capital. 2. Phonemic level- word is analysed as sound. 3. Semantic level – word’s meaning. The more deeply the information is put in the more likely it is to be retained.