Memory

• Memory – the mental process whereby information is encoded, stored and retrieved when requested. It involves three processes: sensory memory, short-term memory and long-term memory. These storage systems differ in term of:

  • Encoding – the process of transforming a sensory input (i.e. sound or visual image) for it to be registered.

  • Storage – the process of holding the information in memory until it is required.

  • Retrieval – the process of locating information that has been stored, and extracting it from memory so that we are consciously aware of it.

[Atkinson and Shiffrin, 1968]

• The first attempt to develop a general model of memory to explain how information flows from one storage to another.

• Components of MSM:

  • Sensory memory - the store that retains the impressions of information received through the senses.

  • Short-term memory - an active memory system which contains all the information currently being thought of.

  • Long-term memory - a store of an infinite amount of information for a long period of time. It is an active system which continuously modifies stored knowledge when new information is presented to it.

:) This was the first model of memory and was influential and useful. It encouraged more research and parts of the model are still used today.

:) It is supported by the serial position effect by Glanzer and Cunitz who found in a free recall test that participants could remember more accurately the words from the start of the list (primary effect) and the words at the end of the list (recency effect) but not the words in the middle.

• Primary effect occurs because the first words that are entered into the STM are being rehearsed and thus transferred successfully to LTM.

• Recency effect occurs because the last words are still in the STM , so were still fresh in STM when the participants were asked to recall the words.

• The words in the middle were recalled most poorly because as more and more words fill the STM, they were not rehearsed enough to be transferred into LTM. Also, by the time they had to recall the words, they would have already faded away.

• This gives evidence for the MSM as it shows that the MSM is made up of separate stores.

:( Has low ecological validity because it was based on laboratory experiments. A laboratory is an artificial setting and therefore does not reflect real life. For example, participants being asked to remember a series of random digits does not reflect how we use our memory in real life. Therefore, caution is needed when generalising laboratory findings in the real world.

:( The MSM states that the transfer of information from STM to LTM is through the process of rehearsal. However, according to other studies this may not be so. For example, Tulving asked participants to read a list of words repeatedly. This words were then included in a longer list which they were asked to recall. The old words which were initially rehearsed were not recalled as frequently as the new words.

:( The MSM has been criticised for being over simplistic. This can be seen through the dual-task experiment of Baddeley and Hitch: participants were told to rehearse a series of 6 digits while simultaneously doing a verbal reasoning task. They found that participants were able to do this accurately. This shows that the capacity of STM is not limited to 7+/-2 as suggested by the MSM, because the digit-span task would have meant that there would be no room left to carry out the verbal reasoning task effectively. It also suggests that the STM may have several different stores which carry out different types of memory functions.

• Demonstrates that the MSM is made up of separate stores:

  • Milner:

  • A young man had brain surgery, as unfortunate consequence of the surgery, memory-related parts of his brain were removed.

  • After the surgery he was unable to form any new memories. He could read the same magazine twice without realizing it. He also forgot people minutes after he had met them.

  • Despite such long-term memory problems, his short-term memory capacity was at a normal level.

  • According to the multi-store model, he was unable to transfer information from short-term to long-term memory, preventing the formation of long-term memories.

  • This demonstrates a distinction between functions of the short-term and long-term memory, therefore supporting the multi-store model.

Glanzer and Cunitz:

• 1) Participants were presented one word at a time and were then asked to recall the words in any order, either immediately (immediate recall group) or after counting backwards for 30 seconds (delayed recall group).

• The immediate recall group - most able to remember the first and last words, the delayed recall group were only able to remember the first words, both groups had difficulty remembering words from the middle of the list

• Both groups could recall words from the beginning of the list because it was possible for both groups to rehearse such first words for long enough so they could be transferred into their long-term memories.

• Words at the end of the list were still being rehearsed and maintained in the STM. Therefore, the participants in the delayed recall group couldn’t rehearse the words in STM, so words were not maintained for recall.

• This demonstrates the existence of STM and LTM and that they are separate stores.

• Craik and Watkins found that participants who were allowed to rehearse certain words were not more likely to remember those words.

  • They argued that elaborative rehearsal may be more important than maintenance rehearsal. Elaborative rehearsal is where the information is elaborated in some way, usually by giving it meaning or by connecting it with knowledge already stored in long-term memory.

  • The MSM model emphasises the effectiveness of maintenance rehearsal, but Craik and Watkins suggest that other methods are more important for transferring information from short-term to long-term memory.

• Brown and Kulik

• Memory for major public events and incidents are called ‘flashbulb memories’.

• These types of memories are when people have a detailed memory of what they were doing and where they were when the event occurred.

• It is argued that there is a brain mechanism that is activated when such emotional events occur, where the scene of learning about the event is printed into memory because of its emotional significance.

• This is evidence against the importance of maintenance rehearsal for memory, where flashbulb memories involve information being transferred straight into long-term memory without rehearsal.

• Devised by Baddeley and Hitch to attempt to explain the structure and function of short-term memory.

  • Consists of four components:

    • central executive

    • phonological loop

    • visuo-spatial sketchpad

    • episodic buffer

• The Central Executive:

  - Is in charge of the other 'slave' systems. Its functions are:

• Allocation: decides which information should receive attention and allocates the information to an appropriate slave system to be processed.

• Control: controls the flow of information to and from the phenological loop, the visuo-spatial sketchpad, episodic buffer and long-term memory and blends the information so our thinking process is combined as one.

• Processes: It can process information in any sensory modality and process different types of cognitive tasks. Evidence suggests information can fade quickly suggesting a limited duration and limited capacity.

• • Phonological Loop: the system that receives, holds and processes sound information. Can be sub-divided:
  • 
    
  • Phonological store (inner ear): a storage system for receiving and holding different types of sound information for a short while.
  • The articulatory control process (inner voice): has two main functions:
    • - Production of sound: converts written information into a sound format. Once a sound is created, this is then deposited in the phonological store.
    • - Rehearsal: allows us to rehearse the sound information held in the phonological store to prevent it from decaying. If the rehearsal is continuous, it will be transferred to LTM.

• Visuo-spatial Sketchpad (inner eye): deals with visual and spatial information (i.e. dimensions). Mental images are created or accessed from the LTM. The visuo-spatial sketchpad can also process verbal or sound information by converting it into a mental image.

• Episodic Buffer: a temporary storage system, which interacts with the other components by manipulating and combining material from the central executive, the phonological loop, the visuo-spatial sketchpad and the LTM to create a single integrated representation.

:) The WMM gives a more detailed explanation of how the STM works than the Multi-store model, shows the STM to be a more complex processing system capable of performing more than one function simultaneously. Whereas MSM suggests STM is a passive single memory store whose main function is to pass information to flow to and from LTM.

:) There is evidence for the phonological loop through the word-length effect experiment. They found that participants can recall more correctly short words that long words. This supports the phonological loop as it shows that capacity is determined by the length of time rather that the amount of items.

:) There is evidence for the visuo-spatial sketchpad. Participants had to follow a visual tracking task and at the same time, had to carry out an imaginary visual task. The participants found carrying out these tasks very difficult because they both required the use of the visuo-spatial sketchpad. However, when the participants performed a verbal reasoning task and a visual tracking task, they found no difficulty performing these tasks at the same time.. This is because they used different slave systems and this supports the WMM because it shows that the STM has different components for different types of memory tasks.

:( Baddeley's word length effect study has been criticised for low internal validity. It is possible that participants would be more familiar with the short words and less so with the long words which may explain why participants found it harder to recall longer words. Therefore caution is needed when generalising laboratory findings to the real world.

:( The WMM only explains the processes of STM. It tells us very little about *** the WMM related to LTM storage of information. This means that the WMM is limited as it only gives us a partial explanation of how memory works.

:( There is still little understanding on the role and function of the central executive, probably the most crucial part of the model. Exactly what it does and how it works still remains speculative. Eysenck (1986) suggested that the central executive might be a single STM store. Again, the model can only give a partial explanation as it cannot explain adequately how the central executive operates.

• Tulving and Thompson (1973) proposed the encoding specificity principle which states that if you are in the same or similar context, i.e. where the information was originally learnt, you are more likely to recall the information because the familiarity acts as a retrieval cue.

• There are two types of cues:

- External cues (context dependent): the environmental surroundings.

- Internal cues (state dependent): is the emotional state the person was in when they first learnt the material. Recall will be better under similar state.

:) There is evidence that external cues can improve memory recall. Godden and Baddeley carried out a field experiment in which underwater SCUBA divers had to learn lists of words either on dry land or under water. They then had to recall the words in the same setting or the opposite setting. They found that words learnt and recalled in the same context were remembered better.

:) Evidence for internal cues by Bower. In an experimental study, participants were hypothesised and imagined a happy or unhappy mood whilst learning information. Participants who recalled in the same mood as they created during learning recalled better than those who recalled in a different mood.

:( Most of the research has been carried out in a laboratory meaning that the conditions are artificial. Asking participants to carry out memory tasks under such conditions may not reveal much in real life usage, suggesting that the findings may lack in ecological validity.

:( Procedural memories – knowing how to do things, are not dependent on cues to aid memory. For example playing an instrument or riding a bike suggests that retrieval cues may only work for some types of memory but not for others.

• Mnemonics:

• Acronym: carried out by using the first letter of each word being remembered to make up a group of words. For example, SCUBA (Self Contained Underwater Breathing Apparatus).

• Acrostic: The first letters of each word are used to make a sentence to be remembered. For example, My Very Educated Mother Just Showed Us Nine Planets (used to remember the names and order of the planets).

• Rhymes: A rhyming mnemonic often uses very similar recurring sounds in words, especially at the end of a verse or sentence. Fore example, I before E except after C. Also, advertising uses rhyme to make sure their product names stick in customers' heads.

:) There is supporting research. Bower tested two groups of students on five lists of 20 words. He found that participants who used the method of loci managed to recall about 72% while the non-mnemonic group managed to recall about 28%.

:( Most of the research findings are based on laboratory experiments under artificial conditions. In real life, abstract concepts and theories are often needed to be learnt, something that mnemonics is not useful for. This suggests that findings may lack ecological validity and therefore cannot be generalised to the real world.

:( Visual mnemonics are less effective on elderly people. Hulick and Grossman found that older people find it difficult to produce and remember visual imagery effectively and are also less inclined to use them as a memory technique.

:( Gruneberg found that one student took up to 20 minutes to make up six letter mnemonic phrase. Park et al found that even experts in the psychology of memory rarely use these techniques in everyday life. This suggests that mnemonics may lack real life application.

• Eyewitness testimony - refers to the accuracy of memory recall of witnesses to specific criminal events often as evidence for investigating crimes in a court of law.

• Loftus et. al carried out a number of experiments to see if giving participants leading questions would affect the accuracy of their memory. 45 students from Washington were shown a video of a car accident. They were then split up into groups and were asked some questions as to what they had seen. One group was given a leading question “About how fast were the cars going when they smashed into each other?” (suggesting that the cars were going quite fast). The other groups were given verbs like hit, bumped, collided, contacted (suggesting that the cars were going at a slower speed). Their answers were collected in a table. The findings showed that the more severe the verbs used in the question, the higher the speed estimation given by the participants. This shows that leading questions can affect the accuracy of recall.

• Loftus carried out a further study to see if misleading post event information influences the accuracy of memory recall. A laboratory experiment was conducted using 150 university students who watched a brief film of a car accident then were asked 10 questions. One of the groups were asked “How fast was the car going when it passed the barn?” which was a leading question as there was no barn. A week later, they were asked more questions, one of which was “Did you see a barn?”. 17.3% replied “yes” whereas only 2.7% from the control group said “yes” showing that introducing false post-event information in the form of a misleading question can actually confuse the original event itself which questions the reliability of EWT

• :) Loftus' laboratory experiments took place in a highly controlled environment. This means that extraneous variables could be controlled or eliminated in order to gain more reliable results.

  :) the study can be easily replicated by other researchers using the same experimental conditions. If similar results are gained, then the study can be seen as reliable and valid.

  :( may lack in ecological validity. Witnessing a video of a crash is different to viewing one in real life. The students were aware that a film was going to be shown to them so they were more attentive than they would probably be in real life when they wouldn't be expecting a situation like that to occur. Therefore, caution is needed when generalising findings.

  :( Loftus' experimental design may have given clues to the participants as to what the study was about, they may have adjusted their answer to fit the expectations of the researcher. These demand characteristics may make the study lack in internal validity.

• Deffenbacher et al carried out two meta-analyses that examined the relationship between anxiety and the accuracy of EWT. In the first meta-analysis, they analysed 27 studies that focused on the relationship between high and low levels of anxiety and face identification. They found that those who experienced high levels of stress identified fewer faces correctly (42%) than those with less stress (54%).

  The second meta-analysis focused on the relationship between anxiety levels and the accuracy of eyewitness recall. In their analysis of 36 studies, they found that those with high levels of anxiety had a correct recall of 52% and those with low anxiety had a correct recall of 64%.

• Johnson and Scott carried out a laboratory experiment to investigate the effect of a weapon on memory recall.. The participants were exposed to two conditions. In the first condition, participants sat outside a room where they overheard a conversation in the next room witnessed a man emerge holding a pen with grease on his hands. In the second condition, participants overheard a heated conversation followed by a man emerging from the room holding a paper knife covered in blood. When the participants were asked to identify the man form a set of 50 photos, 49% were accurate in the no weapon condition compared to 33% in the weapon condition. This shows that anxiety arising from the presence of a weapon distracts attention from other details and therefore reduces the accuracy of recall.

• :( Real life research contradicts laboratory based evidence. Yullie and Cutshall examined a real life crime where 13 people witnessed a stressful event in which an armed thief in a shop was shot dead by the shop owner. Yullie and Cutshall re-interviewed the witnesses 5 months later in which even the witnesses who reported high levels of anxiety managed to recall with about 88% accuracy. This questions the weapon-focus effect as it shows that even in a highly stressful situation, the witnesses' recall of events is still accurate even after a long period of time.

  :( Valentine et al analysed 640 questionnaires completed by eyewitnesses who attempted to identify suspects in 314 line-ups. They found that the presence of a weapon had no effect on the accuracy of correctly identifying the suspect.

  :( Laboratory research is conducted in a highly artificial setting which has low ecological validity. Such artificial setting does not produce the same level of anxiety as a life threatening situation of a real life crime. Therefore, care is needed before generalising for example using the evidence in court cases.

• Leichtman and Ceci carried out an experiment that involved a 2 minute visit in a classroom of pre-school children by a man named Sam Stone. After a visit over a period of 10 weeks the children were interviewed using misleading questions for example “Did you see Sam Stone rip up the book?” although he didn't. They found that 72% of the children claim to have remembered, 44% claimed to have seen Sam do it and even when challenged by the interviewer, 20% still maintained that they had seen it happen. This shows that misleading questions can influence a young child's recall.

• Cohen and Faulkner showed a silent video clip of a kidnapping incident to middle-aged participants and elderly participants with a mean age of 70. Ten minutes later, both groups read an account of the kidnapping which contained misleading information. After a further ten minutes, the participants were asked to fill in a questionnaire about the video clip. They found that the elderly group were more likely to mix up information from the account and the video clip than the younger group. This shows that the elderly are less likely to give an accurate recall as it can easily be distorted by new information.

:( There is evidence that contradicts that the elderly are less likely to give an accurate recall. Valentine et al carried out an experiment in which 3 groups were compared, young children with a mean age of 8, young adults with a mean age of 17 and elderly with a mean age of 70. Participants were shown a video clip of a baby being abducted from a hospital and were then asked some misleading and non-misleading questions. They found that the elderly group were less likely to recall misleading information than the other groups suggesting that the elderly are less suggestible to misleading information.

:( Laboratory research is conducted in a highly artificial setting which has low ecological validity. Such artificial setting does not produce the same level of anxiety as a life threatening situation of a real life crime. Therefore, care is needed before generalising for example using the evidence in court cases.

:( experimental research increases the possibility of demand characteristics which can influence the results. As suggested by Ceci and Bruck, children by their nature seek to please the researcher and give them the answer they think they want to hear rather than their own eyewitness account. This suggest that laboratory findings may be low in internal validity.

• Cognitive Interview technique: Fisher and Geiselman devised the cognitive interview technique to help the police in the process of interviewing to reduce recall error and improve the accuracy of memory recall of the witness.

• Context reinstatement – witnesses are asked to recreate the internal (feelings leading up to the event and after the event) and external (what they saw) context in their mind of the event.

• Report everything – witness is encouraged to report everything they can remember about the event no matter how trivial it may seem.

• Recall in reverse order – the witness is asked to recall backwards in time through the event they witnessed.

• Changed perspective – the witness is asked to recall the event from another witnesses viewpoint.

• Enhanced cognitive interview: Research by Fisher led to the refinement of the cognitive interview. The additions were minimised distractions, encouraging the witness to speak slowly, allowing a pause between the answers and questions and language being adapted to fit the eyewitness' understanding.

:) Geiselman et al produced research evidence for the cognitive interview. They carried out a laboratory experiment by comparing a cognitive interview with a standard police interview. They found that the cognitive interview produced an average of 17% more of the correct information recalled than the standard police interview.

:) To avoid criticisms of the findings being under controlled laboratory conditions which may lack in ecological validity, Fisher et al carried out a real life study. Detectives from the Miami Police department tape-recorded several interviews of victims and eyewitnesses of a crime, some interviews being standard and others being an enhanced cognitive interview. They found that the enhanced cognitive interview extracted 47% more information than the standard one 94% of the recall being accurate. This shows the effectiveness of the enhanced cognitive interview and shows that it has high ecological validity.

:( Cognitive interview does not help with face recognition and person identification. Newlands et al found that the descriptions of criminals were no better using the cognitive interview technique than using a standard interview.

:( Kohnken et al carried out a meta-analysis of 53 studies and found that there was a 33% increase in correct recall using the cognitive interview. However, there was also a increase in recall error with 17% incorrect details reported with the cognitive interview.