Memory

4 groups of participants given a list of words to learn:

• Group A = acoustically similar
• Group B = acoustically dissimilar
• Group C = semantically similar
• Group D = semantically dissimilar

Participants asked to recall the list - some immediately, some after 20 mins

Immediate recall - worst with acoustically similar words

Recall after 20 mins - worst with sematically similar words

Suggests coding in STM is acoustic and LTM is semantic

(-) Didn't use meaningful info

• In real life, semantic coding may be used for STM tasks - generalisability

Invented the digit span technique; researcher reads four digits and increases until the participant cannot remember the order correctly

On average, participants could repeat back 9.3 numbers and 7.3 letters in the correct order immediately after they were presented

(-) Conducted a long time ago

• Lack of control

(+) Results confirmed in other research (eg. Miller)

Made observations of everyday practice 

Noted things come in sevens: 7 notes on the music scale, 7 days of the week, 7 deadly sins etc. 

Concluded the span of STM is 7 +/- 2

Can be improved by chunking

(-) May have overestimated capacity of STM

• Cowan - capacity only 4 chunks

Participants given a trigram to remember and a three digit number

Asked to count back from the three digit number until told to stop - prevent rehearsal

Asked to stop counting after a different retention interval - 3, 6, 9, 12, 15 or 18 seconds

Recall after 3 seconds = 80% accuracy

Recall after 18 seconds = 3% accuracy

Concluded duration of STM is 18 - 30 seconds

(-) Artificial stimulus

• Lacks external validity

(+) We do sometimes try to remember fairly meaningless things

• eg. phone numbers

392 Americans aged 17 - 74

Recognition test: 50 photos from participant's high school yearbook

Free recall test: participants listed the names of their graduating class

Photo recognition decreased from 90% to 70% between 15 and 46 years of graduating - significant about of info still recalled suggesting duration of LTM is potentially unlimited

Free recall less accurate

(+) High external validity

• Lab studies (eg. Shepard) found recall lower when meaningless pictures used

(-) Lack of control of CVs

• Participants may have looked at their yearbook photos and rehearsed memories over the years

Describes how info flows through the memory system

Sensory register - huge capacity but limited duration, info passes into STM if we pay attention

STM - limited capacity and duration, acoustic coding

Info transferred from STM to LTM by maintenance rehearsal

LTM - permanent memory store

Recall info - transferred back from LTM into STM by retrieval

(+) Research support showing difference between STM and LTM

• Baddeley - supports the view that they are separate and independent

(-) May oversimplify STM

• KF - STM for digits read aloud to him was poor, but STM was better when he read them himself - one short term store for visual info and one for auditory info
• WMW can explain this, MSM can't 

(-) May oversimplify LTM

• Research suggests LTM is not a unitary store - one LTM for semantic memories, one for episodic memories 

Episodic memory

• Like a diary
• Memories are time stamped
• Recall is conscious

Semantic memory

• Describe our memories of the world and associated knowledge - like an encycolepia and a dictionary
• eg. knowledge of such things as applying to uni and the meaning of words
• Not time stamped
• Less personal and more about the knowledge that we all share

Procedural memory

• Memories of how we do things eg. riding a bike - 'learned skills'
• Might find hard to explain as recall occurs without awareness or effort - automatic behaviours

(+) Identifying real life LTM stores has real life application

• Belleville et al - episodic memories can be improved in older people with mild cognitive imparements
• Allows specific treatments to be developed

(+) Supported by case study evidence

• HM and Clive Wearing had difficulty recalling events that had happened to them in their pasts, but their semantic memories were unaffected
• One store damaged but other unaffected

(-) Issues with case study evidence

• Lack of control eg. over location of the brain affected or personality variables 
• Difficult to generalise studies to determine the exact nature of LTM

Central executive - attentional process that processes incoming data and and allocates slave systems to tasks, very limited storage capacity

Phonological loop - deals with auditory info and preserves the order in which the info arrives

• Phonological store: stores the words you hear
• Articulatory process: allows maintenance rehearsal

Visuo-spatial sketchpad - stores visual and/or spatial information when required

• Visual cache: stores visual data
• Innder scribe: records arrangement of objects in visual field

Episodic buffer - integrates all types of data processed by the other stores so is decribed as being the storage component of the central executive, as well as linking STM to LTM. Also maintains a sense or time sequencing - recording events that are happening 

(+) Support from case studies

• KF had poor STM for verbal info but could process visual info normally
• Phonological loop had been damaged but other areas were in tact
• However: issues with reliability - unique case with patient who had experienced trauma

(+) Dual task performance studies support the VSS

• Baddeley - participants had more difficulty performing two visual tasks than doing a visual and verbal task at the same time
• Both visual tasks competed for resources

(+) Support from brain scan studies

• Braver et al's participants did tasks involving the CE while having a brain scan
• Activity in the prefrontal cortex increased as task became harder

(-) CE = vague

Interference theory: suggests that forgetting occurs in LTM because we can't get access to memories even though they are avaliable - there is a conflict

Proactive interference: occurs when an older memory disrupts a new one

Retroactive interference: occurs when a newer memory disrups an older one

McGeoch and McDonald - Effects of similarity

• Particpants given a list of words to learn to 100% accuracy, then given a second list which varied in similarity to the original
• Were asked to recall origianl list - performance depended on the nature of the second list - synonyms produced the worst recall
• Shows that interference is worse when the memories are similar - may have been because the new list overwrote the information from the old list (retroactive interference)
• PI also worse when memories are similar - may be because the previously stored into makes new info more difficult to store

(+) Consistent support from lab studies

• eg. McGeoch and McDonald - tightly regulated, control condition
• Increases confidence in validity - objective evidence

(-) Artificial materials in research

• Meaningless world lists do not reflect things we remember in everyday life
• Use of artificial materials makes interference more likely in the lab - may not be a cause of 'everyday' forgetting

(+) Support from real life studies

• Baddeley and Hitch - when rugby players were asked to recall the names of teams they had played that season, it was the number of games they had played that affected recall, not time passed
• Retroactive interference 

Suggests that forgetting occurs when 'cues' present at the time of encoding are not present at retrieval

Cues can be meaningful (directly linked to info) or may be indirectly linked by being encoded at the time of learning (internal state or external context)

Godden and Baddeley - Context dependent forgetting

• Deep sea divers learned list and were later asked to recall them
• 4 conditions: learn on land + recall on land, lean on land + recall under water, learn under water + recall under water, learn under water + recall underwater
• Recall 40% lower when environmental contexts did not match

Carter and Cassaday - State dependent forgetting

• Similar procedure but conditions were drowsy (caused by giving participants antihistamines) or non-drowsy
• Again found that recall was worse when conditions did not match

(+) Everyday application

• Context reinstatement - cognitive interview

(-) Supporting studies lack ecological validitiy

• Contexts are rarely as different as land vs underwater 
• May be best suited to explaining forgetting when the cues are uncommonly distinct - rare in everyday life

(-) Context effects only occur when memory is tested in certain ways 

• Godden and Baddeley replicated their study using recognition instead of recall - no context effect
• Can only explain forgetting for some types of memory, tested in specific ways under certain conditions - poor generalisability

EWT - the ability of people to remember the details of events, eg. crimes or accidents, they have observed

Leading questions

• Response bias explanation: wording of a question does not change the EW's memory, but influences the kind of anser given
• Substitution explanation: wording of a question distorts the original memory
• Loftus and Palmer - participants watched a video of a car accident - estimate of speed dependent on critical verb ('contacted' = nearly 10 mph lower than 'smashed')

Post-event discussion

• Memory contamination - EWs mix (mis)information from others with their own memories
• Memory conformity - EWs go along with eachother because they believe they are right or to win social approval
• Gabbert et al - paired participants watched a video of the same crime, but each from different angles (one could see aspects the other could not). Participants discussed what they had seen before completing an individual interview. 71% of participants recalled info they had not actually seen - gained through PED

(+) Research has real life applications

• Police officers need to be careful about how they phrase questions
• Psychologists can make an important difference to the lives of real people 

(-) Loftus and Palmer's study used artificial materials

• Yuille and Cutshall - 'real' EWs had accurate recall after 4 months
• Artificial tasks tell us little 

(-) Demand characteristics

• Participants might agree to be helpful, when their memories are still accurate
• Challenges the validity of EWT research

Johnson and Scott - Anxiety has a negative effect

• Participants sat in a waiting room and heard an argument in the next room
• Low anxiety condition: man walked through the room carrying a pen with grease on his hands
• High anxiety condition: sound of breaking glass, man walked through the room holding a bloody knife
• 49% could recognise the man in low anxiety condition, 33% in high anxiety condition
• Tunnel theory of memory suggests that the EW's attention is on the weapon (weapon focus) as it is a source of danger and anxiety

Yuille and Cutshall - Anxiety has a positive effect 

• Witnesses to a real life robbery who rated themselves as being more stressed were more accurate (88% compared to 75% for the less stressed group) when interviewed 4-5 months after the incident 

Yerkes and Dodson - Inverted U theory

• Recall increases with anxiety up to an optimum, then accuracy declines (Deffenbacher)

(-) Johnson and Scott's study may have tested surprise rather than anxiety

• Pickel - poor recall in conditions with the highest unusualness eg. raw chicken in a hairdressers 
• Weapon focus effect is due to unusualness rather than anxiety/threat - tells us nothing specifically about the effects of anxiety on EWT

(-) Yuille and Cutshall's study had poor control

• Opportunity for PED and influence of the media 
• Increased recall may not have been because of anxiety - questions validity of research

(-) Inverted U explanation may be too simplistic

• Anxiety has many characteristics - cognitive, behavioural, emotional and physical
• Explanation suggests only physical arousal is linked to accuracy of EWT - fails to account for other factors

Report everything - even seemingly insignificant details may act as a 'cue' (retrieval failure)

Reinstate the context - recalling the weather, location etc. prevents context-dependent forgetting by reminding EWs about the external cues at the time

Change perspective - prevents the influence of schema

Reverse the order - prevents the influence of schema and also makes it harder to lie

Enhanced CI - includes a focus on the social dynamics of the interaction (eg. when to establish and relinquish eye contact). Also includes ideas such as reducing EW's anxiety, minimising distractions, getting the witness to speak slowly and asking open-ended questions

(+) Evidence to support the effectiveness of the enhanced CI

• Köhnken et al - meta analysis shows the enhanced CI consistently provided more correct info than the standard interview

(-) Time consuming

• Getting EW to relax and establishing rapport takes time
• Many forces have only had a few hours training - insufficient
• Despite effectiveness, may be impractical

(-) Increased recall of inaccurate info

• Incorrect info increased by 61%
• Info should be treated with caution