Memory Paper 1

• Coding

  • way information is changed so it can be stored
• capacity


  the amount of information that can be held
• duration


  the length of time a memory lasts
• two models of memory

  1. multi store model
  2. woking memory model
• Who designed the multi store model?

  Atkinson and Shiffrin
• Components of MSM

  1. sensory register
  2. STM
  3. LTM
• What happens with input from SR?

  • if attention is paid to information, it enters the STM
  • if no attention is paid, information fades or decays
• what happens to information in the STM

  • maintenance rehearsal keeps the information in STM
  • elaborative rehearsal puts information in LTM
  • lack of rehearsal means information is forgotten
• SR

  • capacity = very large storage of info from each of senses
  • duration = less than 1/2 a second
  • coding = according to sensory source
• STM

  • capacity = 7 +/- 2 chunks
  • duration = 18 seconds
  • coding = acoustic
• LTM

  • capacity = unlimited
  • duration = a lifetime
  • coding = semantic
• Sperling's research into SR

  • briefly displayed rows of letters and P's could recall 4 or 5 letters, but were aware of more
  • did a further experiment using a partial report procedure
  • used three pitch tones for top, middle, and bottom rows
  • displayed the letters for less than 1/2 a second and found P's recalled on average 75% of the letters in the cued rows
  • This suggests the capacity of the sensory register is very large, but it decays very rapidly, so the duration is very short. 
• Capacity of STM - Jacobs' research

  • measured digit span.
  • He read out lists of digits, starting with 4 digits, asking participants to repeat them back to him. If correctly recalled he repeatied the procedure and increased the number each time, until the participant could not correctly recall the string of digits.
  • the mean digit span for numbers was approximately 9 items and for letters it was 7.
• evaluation of Jacobs

  • very old study meaning it may not have been as carefully controlled as modern research
  • study has been replicated since then and the findings have been confirmed, meaning they are reliable and suggests the study was valid
• Duration of STM - Peterson and Peterson

  • showed P's trigrams and asked to recall it after 3,6,9,12,15 or 18 seconds
  • P's were given an interference task low counting backwards from 100 in 3's to prevent rehearsal
  • After 3 seconds, recall was 80 % and after 18 seconds, recall was only 10 %
  • this supports the idea that the duration of STM is around 18 seconds.  
• Coding in STM - Baddeley

  • investigated the main form of coding in STM by identifying errors in the recall of lists of similar sounding words and similar meaning words.  
  • P's were shown lists of words and asked to recall them
  • When STM was tested, more mistakes came from mixing up words that were acoustically similar rather than words that were semantically similar 
  • p's made more errors with acoustically similar words because they said them to themselves, so similar sounding ones are likely to get confused, supporting the idea that the main form of coding in STM is acoustic.
• evaluation off research into STM

  • research was carried out in highly controlled laboratory conditions meaning that cause and effect relationships could be established
  • standardised procedures means that each research study could easily be replicated by other researchers, allowing the reliability to be assessed
  • The research lacks ecological validity and we cannot be sure that these relationships would apply in everyday memory settings.
• Duration of LTM - Bahrick

  • investigated the duration of LTM by identifying whether p's could remember names and faces of their class mates years after school.  
  • P's were 392 ex-high school students aged 17 - 74, having graduated from high school anywhere from 2 weeks to 57 years ago. 
  • Free recall - free recall the names of their class mates; 15 years after leaving = 60 % accuracy, 48 years after leaving = 30 % accuracy
  • Photo recognition - had to sort photos into those that were or were not in their class; 15 years after leaving = 90 % accuracy, 48 years after leaving = 60 % accuracy 

   
• Evaluation of Bahrick

  • Only one type of long-term memory was being investigated
  • Classmates’ faces might have emotional significance and there will have been opportunity for a great deal of rehearsal given the daily contact classmates will have experienced, and so the findings cannot be generalised to other types of information
  • couldn’t properly control the amount of contact with the yearbook, so this could have affected the results. 
  • Asking p's to recall their class mates tests real life memory
• Coding in LTM - Baddeley

  • aimed to investigate the main form of coding in LTM by identifying errors in the recall of semantically similar/dissimilar words acoustically similar/dissimilar words.
  • P's were presented with one of 4 sets of words and then asked to recall them
  • When p's were asked to correctly recall the word lists after an interval of 20 minutes, p's were most likely to make errors with semantically similar words. 
  • This points to the importance of meaning as words with similar meanings caused confusion, suggesting that information in LTM is coded semantically.  

   
• Evaluation of Baddeley's study

  • lacks ecological validity as it doesn’t resemble the way that memory is used in everyday life.
  • There is good control over the independent variable, and therefore, it is possible to suggest a cause and effect relationship, providing good evidence that coding in LTM is semantic
• evaluation of the MSM

  • Support from studies showing that STM and LTM are different, e.g. Baddeley
  • In everyday life memories are formed related to all sorts of useful things such as people's names and faces, but most of the studies that support the MSM used materials that have very little meaning
  • puts too much emphasis on rehearsal and ignores incidental learning, e.g. gossip
• Evaluation of MSM - Clive Wearing


  • Unable to lay down new LTMs but has some short-term memory, supporting the MSM
  • However, this case study suggests that the long-term memory store is more complicated than just a unitary store
• Evaluation of MSM - HM


  • Unable to form any new long term memories, and he performed well on tests of digit span showing good STM, suggesting these stores are separate
  • However, he was able to develop procedural memories, showing that LTM is not a single unitary store
• Who designed the working memory model of memory?


  Baddeley and Hitch
• Components of WMM

  • Central executive = directs attention to tasks, by allocating them to various subsystems. It has limited capacity, but coding is modality free
  • Phonological loop = limited capacity, acoustically coded, with a brief duration. It is divided into the phonological store and articulatory process
  • Visuo-spatial sketch pad = used to visualise a spatial task, visually coded, capacity is limited to 4 objects
  • Episodic buffer = enables C.E to access LTM and integrate it with other systems. Creates a mental episode and maintains time sequencing and coding is modality free. 

   
• Dual task performance - Baddeley

  • demonstrated how dual task performance is affected by whether the 2 tasks use the same/different components
  • model predicts that it will be harder to do 2 things at the same time if they are both visual tasks/verbal tasks, because both will be using the same store at the same time.
  • However, if tasks differ, they can be done simultaneously as they use 2 stores. 
  • Baddeley gave all participants a visual task, and either another visual task or verbal.
  • Different tasks led to a better performance as the tasks were not in competition for the subsystem's capacity
• Strengths of WMM

  • Support comes from dual task experiments which shows that there must be separate subsystems that process visual and acoustic information 
  • Cohen asked p's to carry out tasks when in an fMRI. When the C.E was working, there was activity in the prefrontal cortex. The occipital lobe of the brain was active when a task was visual, suggesting tasks using different parts of the WM use different parts of the brain
• Weaknesses of WMM

  • The C.E is a very important part of the model, but the model doesn’t give enough information on how it allocates resources
  • EVR had a cerebral tumour removed and performed well on tests requiring reasoning, which suggested that his CE was intact, but he had poor decision-making skills which suggests that his central executive was not wholly intact. This suggests that the CE may in fact be more complex than a single entity as originally proposed
• Types of LTM

  • episodic = memories of events - time stamped, requires conscious effort, not taught
  • semantic = knowledge of facts - not time stamped, requires conscious effort, may be taught
  • procedural = memory for actions - not time stamped, does not require conscious effort, may be taught
• Evidence for types of LTM

  • HM and Clive Wearing
  • it seems that episodic ( perhaps some semantic) memories are affected by damage to the hippocampus and related areas - CW and HM - but procedural memories do not involve the same brain structures. These skill based memories probably involve the motor cortex and the cerebellum
  • Brain scans have shown that episodic memories involve activity on the right side of the prefrontal cortex, whereas semantic memories involve the left side of the prefrontal cortex
• Types of interference

  • Proactive - old interferes with new, as the original memory interferes with the formation or retrieval of new memories which are similar
  • Retroactive - new interferes with old, as the new learning disrupts the retrieval of the older memory associated with it
• Evidence for interference

  • McGeoch and McDonald studied the effect of similarity on retroactive interference. 
  • They asked p's to learn a list of adjectives until they were recalled with 100% accuracy, before learning another list 
  • They were then tested again on the original list, and those whose second list of words were synonyms to the first list did worst, whereas those whose second list of words were numbers did best. 
  • This shows that retroactive interference is greatest with more similar memories. 
• Evaluation of McGeoch and McDonald

  • This has been replicated many times, increasing its reliability and giving greater confidence in this explanation of forgetting.
  • Using lists of words to investigate memory has greater ecological validity than using lists of letters, but is not as realistic as using memory for names and faces, for example, meaning that we cannot be sure that interference is as likely an explanation for forgetting in everyday life, as it is in the laboratory
• Evidence for real life interference - Baddeley and Hitch

  • asked rugby players to recall the teams they had played against that season
  • recall for games played a few weeks ago was better if players had missed one or more games since then, suggesting that interference is the reason that the more regular players performed worse on recall, demonstrating that interference can explain forgetting in real life
  • However the similarity in the names is unclear and there could be confounding variables, particularly individual differences in the memories of those who played more or less games
• Retrieval Failure

  • Some information may be forgotten there are not sufficient cues to retrieve the information we require from LTM, even though the information is stored and therefore available.
  • Cues are triggers that allow us to access information, and they are often coded at the time of learning. They may be internal or external cues. 
  • The encoding specificity principle suggests that cues must be present at time of coding and also at time of retrieval
• context dependent forgetting


  trying and failing to recall information in a different environment to that in which it was learned
• context dependent forgetting - Godden and Baddeley

  • asked divers to learn lists of words either on the beach or whilst they were underwater.  
  • Half of each group were then tested on land, and half underwater.  
  • Divers who learned and were tested in the same environment recalled 40 % more words than those tested in a different environment than the one in which they learned the words
• State dependent forgetting


  trying and failing to recall information in a different internal state to the one we were in when we learned it
• state dependent forgetting - Carter and cassaday

  • participants were asked to learn words either with or without taking a mild sedative beforehand. 
  • Those who learned and were tested in different states did significantly worse on recall than those who learned and were tested in the same state