Memory

• Proactive interference: When previously learned information interferes with new information. Eg. Remembering previous name of a te.acher after she gets married.
• Retroactive interference: Newly acquired information disrupts the recall of old information. Eg. Forgetting your PIN number when you change it to something else
• Retrieval failure:

  When forgetting is caused by an absence of cues. these are signposts that link to memories and help with recall. This has good face validity eg tip of the tongue phenomenon
• Flow of information through the multistore model:

  -Info is encoded in original form in SR-if attention focused on it, it transfers to the stm

  -info coded acoustically in stm-maintained through rehearsal ie repetition

  -adding meaning (elaborative rehearsal) transfers info to ltm

  -lasting memories in ltm have been encoded semantically
• Baddeley and Hitch study on interference:

  tested rugby players memories of names of players they have played against. Players who had played in fewer games recalled more names.
• Gabbert study on post event discussion:60 students and 60 older adults, watched a video of a girl stealing money then discussed. Only half had actually witnessed a crime. 71% recalled information they hadn't witnessed. 60% said she was guilty.
• Lintons diary evidence for capacity of ltm:Linton kept a diary recording daily events, each day = 1 word over several years. After seven years she recalled 11000 events with 70% accuracy. Remember 7/11, capacity research = food
• Johnson and Scott - anxiety reduces recall : Lab experiment, 1 discussion was about equipment failure and a man emerged holding a pen, the other there was shouting and a man emerged holding a knife covered in blood. Condition 1 had 50% accuracy, condition 2 had 33% accuracy. Suggests weapon focus effect.
• Loftus and Palmer - effect of misleading questions: 45 American students were shown a video of a car crash. They completed a questionnaire saying 'how fast were the cars going when they ____ each other?'. When a more violent word filled the gap eg smashed, witnesses estimated that the car was going faster: 41mph (and 32mph for contacted)
• McGeoch and Mcdonald: - interference caused by similar info:

  Participants memorised a list. Then they were split and given either a list of completely different words or synonyms of the first lists. The participants who learned synonyms could only remember 12% of the original list.
• Yuille and cutshall - anxiety reduces recall:21 witnesses to a real shooting were interviewed by police. 13 agreed to be reinterviewed. After 4/5 months errors were rare and accuracy remained high. Misleading info did not effect 10/13 of them.
• HM case study: HM's ability to create procedural memories was unaffected by the removal of his hippocampus. Therefore procedural memories are not created here.
• Fischer: conducted a field experiment with 16 detectives. Groups that received ci training elicited 63% more information
• Golden and Baddeley - retrieval failure:Had divers recall word lists on dry land vs in water. Recall was worse in different contexts.
• Bahricks yearbook study - duration of ltm:Participants had to remember names and faces of high school peers. 90% accuracy after 15 years, 70-80% accuracy after 50 years. Shows life time duration of ltm
• Peterson and peterson - duration of stm:Participants look at nonsense trigrams then count back in 3s and recall after. Time increased by 3s each time. Showed that duration of stm is 18s maximum
• Components of the working memory model

  Central executive: Delegates tasks, manages attention, limited capacity, can code any type of info

  Phonological loop: Slave system, 2 second capacity, auditory information, inner ear holds info, inner voice rehearses info.

  Visuo-spatial sketchpad: Slave system, visual and spatial items, inner scribe plans routes and holds visual info.

  Episodic buffer: Slave system, info in different formats can be combined, communicates with ltm
• Three kinds of ltm:

  Procedural: skills, formed in cerebellum, unavailable for conscious inspection.

  Episodic: memories of events, formed in hippocampus, can be expressed verbally.

  Semantic: memories of info, formed in hippocampus, can be expressed verbally
• Jacobs digit span test - capacity of stm:Participants repeat a string of numbers which gradually increases. Could recall 7 letter and 9 numbers. Chunking can increase capacity
• Baddeley - phonological loop: Participants performed better in recalling lists of short words than long words because they can be silently articulated faster
• Tulving brain scan study:

  Participants perform various tasks whilst being scanned by a brain scanner, left recalled semantic, right recalled episodic
• Baddeley - coding in ltm: same procedure as stm but waited 20 minutes before recall. Semantically similar performed worst so ltm is coded semantically
• Gathercole and Baddeley - separate slave systems:Participants had more difficulty doing 2 visual tasks than 1 visual and 1 auditory. Shows separate stores can work independently
• Baddeley - coding in stm: had people recall lists of 5 words. 4 different ones: acoustically similar/dissimilar, semantically similar/dissimilar.Suggested stm is coded acoustically as acoustically similar performed worst
• 4 main features of the cognitive interview:

  C - context reinstatement: witness returns to place in their mind where the crime happened, helped with context cues

  O - change Order: witness recalls event from end to reverse

  P - change Perspective: describe event from different pov (both help to fill in gaps without schemas)

  E - recall Everything: details act as cues to recall info
• Stores in the multi-store model:

  Sensory register:

  0.5s duration, unlimited capacity

  STM:

  18s duration, 5-9 items

  LTM:

  Lifetime duration, unlimited capacity
• Potentially absent cues:

  Organisation: missing categories that help us group information

  Context: having a different external environment.

  State: being in a different emotional or physical state
• factors that can affect the accuracy of ewt:

  Leading questions: Words in the question activate pre-existing knowledge and expectations ie a schema.

  Post event discussion: Misinformation from other witnesses is combined with their own memory to win social approval or gain information.
• general effect of anxiety on ewt:

  Dodson inverted-U hypothesis: Moderate anxiety improves accuracy of recall, too much causes decline.