Memory

Research into coding

Baddeley

In STM recall, people did worse on words that sound similar

Short Term coding = Acoustic

In LTM recall, people did worse on words that had similar meanings

Long Term coding = Semantic

EVAL: Baddeley study into coding

Artificial stimuli, makes it harder to generalise findings

How is short term memory coded

Acoustically

How is long term memory coded

Semantically

Capacity research

Initial = Jacobs, found mean 7.3 letter recall, 9.3 number recall

Developed by Miller, STM capacity = 7±2

EVAL: Validity of Jacobs study

Jacobs study was done in the late 1800s, not a lot of control in studies

Likely to have confounding variables

EVAL: Millers research in chunks

May have overestimated STM capacity

Cowan : Reviewed other research, STM capacity is 4 chunks

What is the capacity of STM

7 plus or minus 2

STM Duration research

Peterson and Peterson

PPTs given a consonant trigram, and told to count backwards from a number to prevent rehearsal

Each trial was different lengths (3s, 6s, 9s, 12s)

STM Duration = 18-30 seconds

EVAL: Stimuli of Peterson and Peterson

Lacks external validity, consonant trigrams are meaningless stimuli, doesn’t reflect real like recollection

EVAL : Peterson and Peterson limitation

Suggested reason for forgetting things in STM is lack of rehearsal and information is displaced

Peterson and Peterson’s study doesn’t reflect real life conditions of memory

What is STM duration

18-30 seconds

LTM Duration Research

Bahrick

392 PPTs shown yearbook photos

PT1. Recall from a set of 50 photos

PT2. Free recall

15 years out school = 90% recall

48 years out school = 70% recall

EVAL: Validity of Bahrick et al

Studies in LTM with meaningless stimuli show lower recall (Shepard)

By using real-life memories, they increase the external validity

EVAL : Validity of Bahrick et al (NEG)

Confounding variables cannot be controlled in a study including people 48 years out of school

Cannot control when/ how many people have looked at the yearbooks

Who identified multi-store model

Atkinson and Shiffrin

3 main parts of multi store model

Sensory register

Short term memory

Long term memory

Sensory Register : What is it

Intakes information from the environment

5 stores for 5 sense

Main 2 : Iconic and Echoic (sight and sound)

Sensory register : Duration and Capacity

Duration : <0.5 seconds

Capacity : Massive

EVAL: Evidence for MSM

Baddeley: Evidence for STM and LTM coding

Support the idea of separate, independent memory stores

EVAL: More than 1 type of STM

KF : STM recall poor when digits read to him, recall good when he read the digits himself

Highlights there must be stores within the STM

EVAL : More than 1 type of rehearsal

MSM says amount of rehearsal is important to store in LTM

Crank and Watkins = Discovered two types of rehearsal, maintenance and elaborative. Elaborative is needed to move information to LTM

EVAL: Artificial materials

Research studies to support MSM use artificial stimuli (e.g. consonant trigrams)

Who identified the different types of LTM

Tulving in 1985

What are the three types of LTM

Episodic

Semantic

Procedural

Define episodic memory

Recall of events and days

Conscious retrieval e.g. your 17th birthday

Define semantic memory

Knowledge of the world and things, facts

Conscious retrieval e.g. capital of france

Define procedural memory

Memory of how to do things, procedure

Non-deliberate recall e.g. ride a bike

EVAL : Evidence for multiple LTM

HM and Clive Wearing

Semantic and procedural memories unaffected (meaning of words, how to tie shoes, play piano), episodic memories completely non-existent

EVAL: Neuroimaging evidence for LTM

Tulving : PET scanner when performing memory tasks

semantic memories in left prefrontal cortex, episodic memories in right prefrontal cortex

EVAL : Real-life application for multiple LTM types

Belleville et al: Episodic memory can be improved in elderly people with cognitive functions, ptp performed better in memory recall test compared to control

Enables specific treatments to be developed to help develop memory

EVAL: Problems with using clinical evidence

Clive Wearing and HM are good for specific study but lack generalisability

Cannot control different variables

EVAL: 3 types of LTM or 2

Cohen and Squire : Argue episodic and semantic can be ‘declarative memory’, takes conscious effort to be recalled, whereas procedural memory is ‘non declarative memory’

Components in working memory model

central executive

phonological loop

visuo-spatial sketchpad

episodic buffer

Job of central executive

• Is an attention process, so monitors incoming data

• Uses data and makes decision and directs slave systems

Job of phonological loop

• Deals with auditory information and preserves order

• Rehearsal of information

Sub-sections of phonological loop

Phonological store = Holds words

Articulatory process = Allow for rehearsal

Job of visuo-spatial sketchpad

• Stores visual and/or spatial information

• 3 or 4 object capacity (Evidenced by Baddeley)

Sub-sections of visuo-spatial sketchpad

Visual cache = Visual data

Inner scribe = arranges data

Job of episodic buffer

• Integrates info from other stores, maintains order of time

• Links working memory to LTM

EVAL: Clinical evidence for WMM

KF : Poor STM for verbal information, good STM for visual information

Supports the existence of separate visual and acoustic stores in the STM (phonological loop, visuo spatial sketchpad)

EVAL : Dual task performance, WMM

Baddeley: Ptps had more difficulty doing two visual tasks than one verbal and one visual

This is due to the fact that one slave system is being used for two tasks

EVAL : Lack of clarity for central executive

No explanation for the job of the central executive

Needs to be more specific than just ‘attention’

Highlights lack of clarity for WMM overall

EVAL : Word length effects supports phonological loop

Baddeley: Harder to remember long words over short words, due to finite space in articulatory process

Goes away if you do an articulatory suppression task, task that keeps articulatory process busy

EVAL: Bran scan support for WMM & central executive

Braver et al: Tasks involving central executive, found greater activity in left prefrontal cortex

Activity increased as the task got harder

Types of interference

Proactive interference and retroactive interference

Proactive interference

PRevious information disrupts the learning of new information

Retroactive interference

New learning disrupts recall of old information

Similarity effect

McGeoch and McDonald: Ptp had to learn 1 list of words to 100% accuracy, then learn another set of words

Found recall of first list depended on nature of second list, when 2nd list was synonyms of first list, worst recall occurred

EVAL: Lab studies for interference

Heavily researched area, most studies show both types of interference are common ways of forgetting in lTM

Lab experiments have high control and provide valid explanations

EVAL: Artificial materials for studies

Greater chance for interference in lab studies due to artificial materials, lists of meaningless words are likely to be forgotten

EVAL: Real-life studies for interference

Baddeley and Hitch: Ask rugby players to remember names of teams they played in season, as some had missed games their ‘last’ team had been 3 weeks prior

Recalls was more accurate when player had played less teams in between

Interference can be applied to everyday situations

EVAL: Time for learning in lab studies

Time in between learning and recalling lists is often shorter than what would be found in real life situations

EVAL: Interference can be overcome using cues

Tulving and Psotka: Gave ptps 5 lists of 24 words, with 6 categories (not stated)

Recall was 70% for first list, fell with each list

At end did cued recall of first list, told names of categories, and recall rose back to 70%

Main reasons for retrieval failure

Encoding specificity principle

Contest- dependant forgetting

State-dependant forgetting

Encoding specificity principle

Discovered by Tulving

If a cue is to be helpful to recall, must be there at learning and at recall

If cue is there at learning but not at recall, forgetting can be expected

• Mnemonics, music, taste, sound

Context-dependant forgetting

Context around person matches up when learning and recalling e.g. environment

Context-dependant forgetting study

Godden and Baddeley

18 PPTs learning lists of 38 words

When conditions matched up (land or underwater), there was better recall

State-dependant forgetting

Similar to context dependant, but is more about the internal state of the PPT

State dependant forgetting study

Carter and Cassidy

Gave PPTs anti-histamines, made them drowsy

Learnt words drowsy and not drowsy

When mismatch in conditions, recall went down

EVAL: Supporting evidence for retrieval failure

Godden and Baddeley, Carter and Cassaday are two main examples of research

Eysenck: Retrieval failure is main reason for forgetting from LTM

EVAL: Questioning context effects

Baddeley: Context effects are not as strong as studies show, contexts have to be very different in order to have effect

Land V.S water are stark contrasts, so difference is potentially exaggerated

EVAL: Recall vs recognition for Godden and Baddeley

Godden and Baddeley: Replicated study but with recognition test, no context-dependant effect, performance same in 4 categories

Limitation = Presence or absence of cues depends on how they are being tested

EVAL : Testablity of ESP

No way to test the ESP with 100% surety, we make assumptions based on outcomes of tests

No way to independently establish whether or not cue has been encoded

EVAL: Real-life applications

Though the effect may not be massive, they are visible in real life E.g. going into a room for something, forgetting but when you go into the original room you remembered, it will come back

Basic principle of cognitive interview

EYEW TESTIMONY: Leading Q study

Loftus and Palmer

PPTs watch a video of a car crash and are asked how fast the car was going when it __?

Changing of the verb changed the answer that was given

• Contacted = Mean of 32mph

• Smashed = Mean of 40.5mph

EYEW TESTIMONY: Leading Q: Explanations

Response bias: Wording of question influences how a person makes their decision

Substitution : Wording alters memory

EYEW TESTIMONY: Post Event Discu: Study

Gabbert et al.

Watch videos of a crime from different perspectives, PPTs studied in pairs and talked before they did a test on what they saw

Control group = 0% mistaken info

PPTs group = 71% mistaken info

EYEW TEST

Post Event Discuss : Explanation

Memory conformity = Go along with each others perceptions

(NSI/ISI?)

EVAL: Real life application of eyewitness testimony

Lotus: Found leading questions can distort memory when police officers interview witnesses

EVAL: Artifical test in Loftus and Palmer

Watching clips is different to real incident, lacks the real emotion of the event, which impacts recall

EVAL: Individual differences in recall

Older people less reliable than young people, Anastasi and Rhodes: 18-25 and 35-45 were more accurate than 55-78

All age groups more accurate in identifying their own age

EVAL: Demand characteristics

Answers in lab studies of EWT are result of demand characteristics, want to appear helpful to research

Easy if the questions are yes/no

EYEW TEST

Anxiety → negative study

Johnson and Scott

Johnson and Scott

Procedure

Led PPTs to a waiting room and they hear an argument from next room

High anxiety = Hear glass smash, see man walk out with knife covered in blood

Low anxiety = Man walks out with pen and grease on hands

Johnson and Scott

Findings

PPTS asked to identify man out of 50 pictures

High Anxiety = 33% correct identification

Low anxiety = 49% correct identification

What theory helps explain Johnson and Scott findings

Tunnel theory

• Witness’s attention is drawn to the weapon as it is the source of anxiety

EYEW TEST

Anxiety → Positive study

Yuille and Cutshall

Yuille and Cutshall

Procedure

13 witnesses of shooting and robbery in Vancouver

Interviewed 4-5 months after incident

Asked to recall details and rate how stressed they were + any emotional issues since

Yuille and Cutshall

Little change in details in 5 months since police interview

High anxiety = 88% accurate recall

Low anxiety = 75% accurate recall

Yerkes-Dodson Law

Also known as inverted-u theory

• There is an optimum level of anxiety

• Too high is just as bad as too low

EYEW TEST

Additional studies for anxiety

Parker et al.

Valentine and Mesout

EYEW TEST

Parker et al.

Hurricane Andrew, interviewed people and judged anxiety by amount of damage done on house

• Medium amount of anxiety = optimum recall

EYEW TEST

Eval Parker et al

Is damage done to property the best way to operationalise anxiety levels

EYEW TEST

Valentine and Mesout

London Dungeons, reduced entrance fee to wear a heart rate monitor

High anxiety PPTS = less accurate details when asked to describe an actor they had encountered

• High anxiety = 17% identified the actor in line-up

• Low anxiety = 75% identified actor in line-up

EVAL: Weapon focus may not be relevant

Pickle: Did a study at hairdressers, using scissors, handgun, wallet, raw chicken, objects ranked on unusualness for the scenario

Eyewitness accuracy was lower for the more unusual objects (handgun and chicken)

EVAL: Field studies lack control

Things happen in field that researchers have no control over (media, external discussion of the event)

Extraneous variables maybe be responsible for accuracy of results

EVAL : Ethical issues in inducing anxiety

E.g. Johnson and Scott, idea that someone has been injured

Potentially unethical as it may create psychological harm

EVAL: Inverted U theory is too simplistic

Anxiety is hard to define and measure, inverted U theory assume only physiological arousal is linked to poor performance

EYEW TEST

4 aspects of cognitive interview (CI)

Report everything

Mental reinstatement

Change perspective

Change order

Rude Men Can’t Cum

EYEW TEST CI

Report everything

Include every detail even if it doesn’t feel relevant

• Trivial details may trigger other important details

EYEW TEST CI

Mental Reinstatement

Such as returning to scene in mind and imagine the circumstances

• Weather, sensory input

  • Context dependant forgetting (Godden and Baddeley)

EYEW TEST CI

Change order

Recall in reverse chronological order or start during middle of event

• Prevents people from reporting what they EXPECT to happen

• Harder to be untruthful

EYEW TEST CI

Change perspective

Imagine themselves in the perspective of another witness or the perpetrator

• Disrupts expectations and schema of the crime

EYEW TEST

Enhanced Cognitive Interview

Forms a deeper relationship between interviewer and interviewee

Witness can talk freely before asked questions

• Encourages open recall

Modified Cognitive Interview

Removes the C’s from RMCC

Suitable for children, stresses relationship and allows witness to control narrative

Studies for CIs

Geiselman et al - CI is more accurate than SPI

Fisher and Geiselman - Amended the CI to form better relationship, ECI

Meta-analysis for CI

Kohnken et al

Memon, Meissner, Fraser

Kohnken et al

• Combined information from over 50 studies

• Found CI’s produce 81% correct info

  • Produces more incorrect info - 61%

Memon, Meissner, Fraser

• Comparing CI to other methods (ECI, MCI)

• CI produced more accurate detail, more than in Kohnken

  • High amounts of confabulation in MCI