Memory

Coding research and evaluation

Coding- format in which information is stored in various memory stores

Baddeley- gave different lists of words to 4 groups of participants, and had to recall them in correct order

• Immediately (STM)- best with acousticaly dissimilar: STM coded acoustically

• 20 mins (LTM)- best with semantically dissimilar: LTM coded semantically

G1 (acoustically similar)- cat, cab, can

G2 (acoustically dissimilar)- pit, few, cow

G3 (semantically similar)- great, large, big

G4 (semantically dissimilar)- good, huge, hot

Separate memory stores

• identified clear difference between 2 memory stores

• STM being acoustic and LTM being semantic has stood the test of time

• helped lead to multi-store model- some exceptions though

Artificial stimuli means there is limited application

• word lists do not have personal meaning

• means findings do not tell us much about coding in different memory tasks in everyday life so people may use semantic coding even for STM

Capacity research and evaluation

Capacity- amount of information that can be held in a memory store

Jacobs- researcher reads out 4 digits and participants repeats them back, if correct then the research ups amount of digits read- found capacity was 9

Valid study as has been replicated- despite being old and lacking adequate controls, more controlled recent studies have found the same findings

Miller- made observations of everyday practice and found that lots of things come in sevens e.g. days, notes- though capacity was 7± 2

• people can read 5 words as easily as 5 letters through chunking- grouping digits or letters into chunks

May have overestimated STM capacity- Cowan reviewed other research and concluded STM is 4 ± 1

Duration research and evaluation

Duration- length of time information can be held in memory

Peterson and Peterson (24 students have 8 tests each)

• given consonant syllable e.g. YCG then told to count back from a 3 digit number (stops maintenance rehearsal)

• different tests told to stop and recall between 3-18s→ 3s=80%, 18s=3%

• STM duration approx. 18s unless we do maintenance rehearsal

Stimulus was artificial- does not reflect everyday activities where we try to remember something meaningful so lacks external validity

Bahrick- 392 Americans aged 17-74 using school yearbooks to test recall

1. remembering names from 50 photos- +15 years= 90%, +48 years=70%

2. free recall to just remember names +15 years= 90%, +48 years= 30%

shows LTM may last up to a lifetime

High external validity- researcher investigated meaningful memorie so findings are a real estimate on duration

Types of long term memory, tulving

Episodic memory: ability to recall information from our lives

• time stamped as to when they happened and how they timely relate to other events

• include several elements e.g. people and place

• conscious effort to recall

Semantic memory: shared knowledge of the world e.g. word meanings

• less vulnerable to distortion and forgetting

• conscious effort to recall

Procedural memory: for actions and skills

• may find hard to explain to someone

• no conscious effort to recaoo

Explanations for forgetting- interference

Two pieces of information disrupt each other- forgetting or distortion

In LTM these memories may be available but inacessible

Proactive interference: older memory interferes with new e.g. teacher learnt many past names so cannot remember new names

Retroactive interference- newer memory interferes with older memory e.g. can no longer remember old memory of phone number as have new one

McGeoch and McDonald: studied retroactive interference by changing similarity between 2 sets of material.

• Participants learnt 1 list of 10 to 100% accuracy

• Then learnt new list to 100% accuracy: synonyms, antonyms, unrelated, consonant syllables, 3 numbers, control

• Then had to recall original list, those who had learnt synonyms had the worst recall because the new information was interfering with the old

• Shows interference is worse when memories are similar

Evaluation for long term memory types

Clinical evidence from HM and Clive Wearing

• HM had episodic memory impaired due to brain damage from operation e.g. could not recall stroking dog 30mins earlier but could remember concept of a dog (semantic)

• Clive Wearing had episodic memory impaired by viral infection but could stil read music, sing and play piano (procedural)

• different memory stores can be damaged but others not

→ clinical studies lack control of variables as the researcher did not study memory before

Real world application- allows psychologists to help people with memory problems

• age leads to people forgetting recent episodic memories

• Belleville created intervention to improve episodic memory in the elderly- trained participants did better on episodic memory testing than a control group

Conflicting neuroimaging evidence (no clear support0

• Buckner and Peterson: semantic memory is left prefrontal cortex and episodic memory is on the right

• Tulving: argues episodic memory is left prefrontal cortex and episodic memory is on the right

Cohen and Squire: argue episodic and semantic memories are stored together in one memory store as both are consciouslly recalled- ‘declarative memory’

Tulving has argued episodic memory is sub category of semantic memory (most memories are a combination of both)

• it is possible to have a functioning semantic and damaged episodic BUT NOT damaged semantic and functioning semantic

• Hodges and Patterson: some people with Alzeimers can form new episoidic memory but not semantic

Evaluation for the multi-store model

Research shows that LTM and STM are different and independent

Baddeley- mix up words that sound similar in STM, and words with similar meanings in LTM

→ shown using artificial stimuli so MSM not valid model of how memory works daily

Serial position curve shows that there are separate stores

• primacy: words that appear first in word list are more likely to be recalled as they are rehearsed

• recency: words at end of word will still be in STM

More than one STM store

• KF had amnesia due to a motorbike accident- STM poor when read aloud to but better when he read to himself (separate store for non verbal sounds)

Prolonged rehearsal not needed to transfer to LTM

• MSM argues the more we rehearse something the more likely it is to go to LTM

• Craik and Watkins argue that the type of rehearsal matters more- elaborative rehearsal is needed: linking rehearsal to existing knowledge or thinking about meaning

Flashbulb memories are detailed, vivid memories of event stored immediately and lasts forever e.g. 9/11, wedding day

• often life changing so emotionally aroused when memory first encoded

• not everything is rehearsed to go to LTM

Evaluation for the working memory model

Clinical evidence from KF- after brain injury, KF had poor STM for auditory but could process visual information. His recall of letters and digits better when he read them than when they were read to him

→ phonological loop was damaged but visual spatial sketchpad still intact (supports model)

Studies of dual task performance

• Baddeley’s participants could do visual and verbal task at the same time

• performance declined when both tasks were visua or both verbal

• similar tasks compete for same slave subsystem but not compeition when they are different tasks

→ used tasks unlike tasks that we perform in our everyday lives and done in lab conditions

Nature of the central executive- lack of clarity over how the central executive works

Baddeley “most important but least understood part of the working memory”

• needs to be more specified as being unclear challenges the integrity of the model

Evaluation for retrieval failure

Retrieval cues can help overcome forgetting in everyday life- may not have a strong effect but worth paying attention to e.g. environment where we first learnt

Research support- lack of relevant cues at recall can lead to context and state dependent forgetting, occurs in real world situations

→ Baddley argues that context effects are not very strong especially in everyday life as environments are generally not different enough so does not explain much everyday forgetting

Context effects may depend substandially on the type of memory being tested

• Godden and Baddeley replicated underwater experiment but used a recogniton test insetad of recall- had to say whether they recognised words read to them from a list

• Performance was then the same in all 4 conditions showing context effects only apply when a person has to recall

Encoding specificity principle- how is it possible to know when a cue is encoded, means certain studies are based on assumptions rather than actual evidence

Evaluation for cognitive interview

Kohnken- supporting evidence

• did meta analysis of 55 studies comparing cogntive interview with standard police interview

• cognitive interview gave 41% higher accurate information compared with standard review (only 4 in analysis showed no difference)

• shows cognitive interview helps get information not immediately accessible

→ enhanced cognitive interview found more incorrect details (quantity>quality)

Reinstate the context and report everything prevents issues with leading questions and context-dependent forgetting

Not all elements of cognitive interview are useful, Mihne and Bill- some find reverse the order and change perspective confusing which weakens credibility of overall cognitive interview

Cogntitive interview is time consuimg

• completing it takes more time and training than the standard police interview

• more time to build rapport and allow witnesses to relax

• not realistic for police forces to do

Evaluation of anxiety affecting eye witness testimony

Valentine and Mesout- research evidence

• volunteers wore heart monitors in horror labyrinth in London dungeons

• split to high and low anxiety

• had to describe person they encountered in labyrinth- those with low anxiety had 75% recall, those with high anxiety had 17% recall

• anxiety has a negative effect on recall

Christianson and Hubinette- research evidence

• interviewed 58 witnesses to actual bank robbery (some bystanders and some workers)

• 75% overall accuracy- workers more direct so therefore more anxious had higher accuracy

• anxiety has a positive effect on recall

→ interviewed them 4-15 months after, other factors such as post event discussion may be a confounding variable

Inverted-U explanation is too simplistic

• anxiety now hard to define and measure accurately as has many characteristics e.g. cognitive, behavioural, emotional

• ignores other aspects that impact eye witness testimony

Johnson and Scott, pen vs knife- participants may focus on weapon not person because it is unusual not because it is anxiety inducing

• Pickel showed participants video of hairdressing salon where person had gun, scissors, raw chicken or wallet

• eye wtiness testimony for actual person lowest when holding chicken or gun as these were the most unusual

• shows weapon focus is due to unusualness so does not shows anxiety’s effect on EWT.

Evaluation for misleading information effecting eye witness testimony

Research into misleading information has practical uses in the criminal justice system

• consequences of inaccurate eye witness testimony can be serious so police have to be careful

• psychologists can act as expert witnesses and improve the way the legal system works in order to protect innocent people

→ practical application of eye witness testimony of eye witness testimony affected by research issues

• loftus and palmer (car having bumped, collided etc.) watched their film clips in a lab which is different from the real event

• eyewitnesses have more important consequences in real world as less motivation in research

• eye witness testimonies may be more dependable than studies suggest

EWT is more accurate for some aspects of an event than others

• Sutherland and Hayne, participants had more accurate recall for central details because core memories are resistant to misleading information

Demand characteristics- answers given by participants in lab studies are because participants want to be useful and not let researchers down (often guess)

Evidence that post event dicussion alters eyewitness testimony

• Skaberg and Wright, film clips where one mugger had light brown hair and the other dark

• PED led to a blend of two “medium brown”- memory distorted by contamination

Evaluation for interference

Real world interference- increase validity of theory

• baddeley and hitch asked rugby players to recall the names of teams they had played against during a season.

• players who had played the most had the worst recall because they had the most interference

→ interference is unusual as conditons necessary are rare as memories have to be fairly similar to interfere with each other- occasionally possible but not often

McGeoch and McDonald and other studies involve learning artifical words in a lab setting so less motivation, does not reflect how we usually use memory (lacks ecological validiity)

Interference is temporary and can be overcome using cues

• Tulving and Psotka gave participants list of words organised into categories. 70% recall for first list but worsened as participants learnt each additional list. when given a cue for the category, recall rose back to 70%

→ interference caused temporary loss of accessibiity which is not predicted in interference theory.

Anxiety affecting eye witness testimony

Negative effect on recall- anxiety creates physiological arousal which prevents us from noticing important cues (presence of a weapon creates anxiety)

Johnson and Scott- participants believed they were doing a lab study and waited in waiting room

Condition 1: heard casual conversation and then man came out with pen

Condition 2: heard heated argument, broken and glass and then man came out with bloody knife

• 49% of C1 could identify man out of 50 photos

• 33% of C2

tunnel theory- vision focuses on weapon not man

Positive effect on recall- anxiety triggers fight or flight which increases alertness and memory, from an evolutionary perspective this is adaptive so that you know how act in future danger

Yuille and Curshall- studied actual shooting in gun shop where the owner shot the thief dead

• 13 witnesses took part 4-5 months after

• measured on details they recorded and stress they felt at the time on scale of 1-7

• highest stress 13% more accurate than least stressed (anxiety may enhance EWT)

Deffenbacher- reviewed 21 studies of EWT, and argued there is an optimum level of anxiety which helps us to have balance between fight or flight, an pure panic- Yerkes/Dodson law

Cognitive interview improving eye witness testimony

method of investigating/ interviewing eyewitnesses to help them retrieve more accurate memories- obtains maximum information while minimising interviewer’s influence

Fischer and Geiselman

1. Report everything: include every detail of event as may be important as it triggers more important memories.

2. Reinstate the context: witness should return to crime scene in their mind and imagine environment and emotions- stops context dependent forgtting

3. Reverse the order: events recalled in different order to original sequence to prevent people reporting expectations, also harder to lie

4. Change perspective: recall as if from other person’s perspecive (stops schema)

enhanced cognitive interview- focuses on social dynamics e.g. reducing anxiety, slowing speech, minimising distractions, open ended questions, no judgement, adapting language

explanation for forgetting- retrieval failure

encoding specificity principle- cue has to be present at encoding/ learning and at retrieval. If cues are different/ absent at retrieval than encoding there will be some forgetting (not all cues are meaningful)

Context dependent forgetting- depends on external cue

• Godden and Baddeley: studied deep sea divers who learnt list of words then recalled

1. land/ underwater

2. land/ land

3. underwater/ land

4. underwater/ underwater

accurate recall was 40% lower in non matched conditions as external cues available at learning were different from at recall- retrieval failure.

State dependent forgetting- depends on internal cue

• Carter and Cassaday: gave antihistamine drugs to participants leading to drowsiness

1. on drug/ on drug

2. not on drug/ on drug

3. not on drug/ not on drug

4. on drug/ not on drug

learnt list of words- memory worse when state at learning and recall were different

Misleading information affecting accuracy of eye witness testimony

Post event discussion- when witnesses discuss what they have seen with cowitnesses and other people which influences accuracy of witness’ recall

Gabbert- in pairs participants watched video of same crime but from different POVs

• 71% mistakenly recalled aspects of events they did not see but had picked up in dicussion

• 0% of control group (no discussion) made mistakes

Memory contamination- memories became altered and combined in discussion

Memory conformity- witnesses go along with each other to win social approval or because they doubt themselves (actual memory unchanged)

Leading questions- questions phrased to suggest a certain answer

Loftus and Palmer- 45 students watched film clips of car accidents and asked “how fast were the cars going when they … each other”

smashed, collided, bumped, hit, contacted

(40.5mph—31.8mph)

response bias: wording has no effect on memory just influences how they answer

substitution explanation: wording of leading question does change participant’s memory

• those who heard “smashed” more likely to say there was broken glass

the multi-store model

maintenance rehearsal: repeat material to ourselves to stay in STM

prolonged rehearsal: when we rehearse information enough to pass to LTM

Retrieval: when we want to recall information from LTM it has to pass to STM

Sensory register- all stimuli from environment pass into this

• coding= depends on store e.g. iconic= visual, echoic= sound

• duration: less than half a second

• capacity= very high (each eye has 100 million cells)

Short term memory

• coding= mainly acoustic

• duration: 18s unless rehearsed

• capacity= 7± 2

Long term memory

• coding= mainly semantic

• duration= lifetime

• capacity= practically unlimited

the working memory model

concerned with mental space that is active when we manipulate and temporarily store information

central executive- monitors incoming data, divides attention to different slave systems

• limited processing capacity

• does NOT store information

visuospatial sketchpad- stores visual and spatial information

• capacity: 3-4

• visual cache= visual data

• inner scribe= arrangement of objects

phonological loop: auditory information preserved in order it arrives

phonological store: stores words you heard

• coding: acoustic

• capacity: 2s we read

articulatory process: does maintenance rehearsal so repeats words in loop to keep in working memory model

Episodic buffer- merges visual, spatial and verbal information with time sequencing into the long term memory (storage component of central executive)