Memory

Gree Amb Re
Personalised Learning Checklist: Memory
n er d
Describe the multi-store model of memory and its components:
sensory register, short-term memory and long-term memory. /
Know the features of each store: coding, capacity and duration. /
Discuss the three types of long-term memory: episodic, semantic,
procedural. /
Describe the working memory model: central executive,
phonological loop, visuo-spatial sketchpad and episodic buffer. /
Know the Features of the working memory model: coding and
/
capacity.
Discuss the explanations for forgetting: proactive and retroactive
interference /
Discuss the explanations for forgetting: retrieval failure due to
absence of cues. /
Describe the factors affecting the accuracy of eyewitness testimony:
misleading information, including leading questions and post-event
discussion; anxiety. /
Discuss how to improve the accuracy of eyewitness testimony,
including the use of the cognitive interview. /

Memory: The Multi Store Model of memory

- Atkinson and Shiffrin (1968)

- Information processing approach from cognitive approach
- One of the earliest models of memory
- Linear model passes info from 1 store to the next + memories are
formed sequentially.
- Sensory store- 250 miliseconds
- Each store is unitary. This means that each store is one system.
There are no sub systems or further divisions (unlike other models
of memory)
 Encodi Way info is changed so it can be stored in the memory, 3
ng main ways in which info can be encoded: visual (picture),
acoustic (sound), semantic (meaning)
Capacit How much info can be stored (7+- pieces for STM and
y unlimited for LTM)
Duratio Period of time info can last in the memory
n

Strengths Weaknesses
Pioneering model of Low ecological validity for some research- stimuli p’s asked to
memory that inspired remember bear little resemblance to items learned in real life
further research + other (Peterson + Peterson’s trigrams)
inf models like WMM Over simplified- evidence suggests multiple short + long-term
Large base of research memory stores
that supports distinct STM Does not make much sense to view procedural memory (type
+ LTM systems (patient of LTM) as being encoded semantically i.e. knowing how to ride
KF’s impaired STM a bike thru meaning.
following a motorcycle Assumed that LTM has an unlimited capacity, as research has
accident, LTM remained been unable to measure this accurately.
intact)

Experiments Case Studies

Miller (1956) supports that STM has a Patient HM- supports that STM + LTM are
capacity of 7+/- 2 pieces of info different stores, and neuroscience evidence to
Peterson + Peterson (1959) support the suggest memory is in the hippocampus
limited duration (20-30 secs) of STM Patient KF- refutes that the STM store is a
Bahrick (1975) supports the unlimited unitary store. Suggests there are diff types of
duration of LTM STM.
Patient Clive Wearing- unable to transfer info
from LTM
Miller (1956)/ Jacobs (1887)

- Miller cited for research examining the STM, Jacobs conducted

similar research 69 years ago
- Jacobs examined capacity of STM using a digit span test, on average
p’s could recall 7.3 letters + 9.3 numbers
- Primary and Recency effect- remembering the first and last parts of
a sequences

Duration of LTM- Bahrick et al. (1975)

- Wanted to investigate duration of LTM, examined 392 American uni

grads on their memory for former classmates by asking participants
to match the names of former classmates to pictures in a yearbook
photo
- 90% accuracy after 14 years 60% accuracy for matching names +
faces 47 years after graduation
- Memories stored in LTM can last a life-time or at least 47 years,
supports multi-store model- our LTM has a lifetime duration +
semantically coded

Trigram

Peterson + Peterson (1959)

◦ Lab experiment - examined the effect of duration on STM

◦ 24 Psychology students were shown 3
letter words, followed by a distractor
task of 3-18 seconds.
◦ The distractor task prevented the
rehearsal of the letters.
◦ They then tried to recall the words.
◦ After 3 seconds, 80% of the words
were recalled, whereas 10% were
recalled at 18 seconds.
 ◦ This shows that STM has a limited duration (approx. 20-30 seconds)
◦ This shows that rehearsal is vital for memories to move into LTM
◦ Therefore, supports the MSM
Evaluation

 Lab = good controls i.e. length of intervals and same conditions for
each Participant, makes procedure easy to replicate and reproduce
data (scientific)
- Lab = low ecological validity task not reflective of a real life
memory task- don’t memorise trigrams in everyday experience.
- Lab = possible demand characteristics – P’s could have figured
out the task and therefore confounded data through behavioural
changes
- Lab = possible deception to P’s to try an eliminate demand
characteristics
- Sample = were psychology students, so could have known about
memory strategies or guessed the aim more easily
Case Studies

◦ Patient HM- 8cm of hippocampus removed in temp. lobe, supports

multi-store model that STM + LTM are separate. Neuroscientific
evidence
◦ Patient KF- injured in a motorcycle accident, could recall LTM but not
STM- supports MSM model, could remember visual info but not
sound- disputes MSM. Has to be 2 storage systems- auditory
memory lost but visual component retained
◦ Patient Clive Wearing developed a herpes virus that attacked his
CNS, could not form any new LTM, STM intact. Supports duration of
STM, transference + rehearsal is needed for LTM storage.
Neuroscientific evidence.

The Working Memory Model

- Developed by British psychologists Baddeley and Hitch in 1974

- They built upon MSM to challenge the idea that STM is a unitary
store

Main assumptions:

- STM is active + complex. Should be called a working memory

- Working memory is a system which holds info in your mind as you
are doing something with it
- There are multiple systems for different types of info: Central
executive, Episodic buffer, Phonological loop, Visuo-spatial
sketchpad
 - There is a limited capacity to Working Memory (STM) and it is a
temporary store of info
- Any info that is not used is lost/forgotten i.e. distraction,
overloading, overworking

Dual-Task Paradigm

- Explains why + how we can do 2 diff tasks at one time i.e. drive a
car (visual) whilst talking to a passenger (auditory) but we cannot
do two tasks together i.e. listen to music (auditory) + have a convo
(auditory)
- P’s had to do an interference task. There was a control group and it
was predicted that the visual group would perform worse. This is
because the visuo-spatial sketchpad was overloaded

Alzheimer's Disease + role of central executive

- Evidence for the co-ordination of the central executive shows a

decrease in function with Alzheimer's
- Patients had an impairment in completing tasks that required all
components of the sub-system but could complete individual single
system task (visual)
- Evidence that CE is needed to co-ordinate sub-systems

Patient KF

- Suffered STM problems following a motorbike accident where he

damaged his temporal lobe. Visual memory intact but his
phonological memory was impaired

Research into separate visual + spatial systems

- Evidence for the separate comps. Of the visuospatial sketchpad

- Spatial inference task interferes with spatial memory but not with
visual memory

Experimental evidence

- Results from dual-task exp. show that there must be diff systems for
processing diff modalities of sensory info
- Evidence that distinct sensory sub systems exist in STM

Neuroimaging

- Evidence for localisation of diff sub-comps of working memory

- PET scans reveal that Broca’s area is responsible for lang +
phonological stores
- Hard to localise specific brain areas for specific functions- seen to be
diffused across whole cortex (brain surface)
 Evaluation of WMM
- Patient KF supports the WMM (injured in motorcycle accident, lost
his ability to recall sounds in his STM. His visual STM was unharmed
– therefore supports the WMM proposal that STM is made up of
multiple slave systems and refute the MSM that STM is a unitary
store.
o however, case study evidence may not be generalisable, KF
had brain trauma so doesn’t represent normal brain
functioning
o But idiographic research like KF is vital to support nomothetic
theories like WMM (and refute MSM)
- Dual task studies by Baddeley and Hitch (1976) support the concept
of a working STM and explain how we can perform two STM tasks at
the same time if they come from two different slave systems.
o Therefore, is evidence for two slave systems (v-s sketchpad
and phono-loop)
o But evidence to support uses artificial tasks which are low in
ecological validity so don’t really reflect real life cognitive
(thinking, processing and memory) functions and could be
considered reductionist.
- The transference of STM to LTM is not explained very well in the
WMM and could therefore be an incomplete model.
o However, the episodic buffer was added to try and explain
this, and it remains one of the more complex model of STM
and inspired lots of future research within memory research.
- The WMM is much better at explaining the STM as a collection of
active processes rather than a single unitary store that just
rehearses information (MSM)
- Lack of detail on the role of the central executive
Central Boss- supervises the system + function to control flow of info. Can move btw
Executive tasks by controlling + selecting attention. Supervisory activating system
(SAS) that controls cognitive processes (not just memory). Switches attention
+ cognitions btw STM + LTM. Directs info to the 2 slave systems. Works in
any sensory modality (an 5 senses).
Visuo- Slave system- a temp storage system + holds visual + spatial info +
spatial function links to non-verbal intelligence. In the right hemisphere of the
sketchpad brain: 1st part- visual cache- stores colour + shape info. 2nd part- ‘inner
scribe’- deals w retrieval + rehearsal of mental images.
Episodic Added in 2000, provides time sequencing for visual, spatial + verbal info (ie
buffer using chronological ordering). Brings info from LTM to help w STM function
(i.e. rules for chunking + sequencing). The ‘storage space’ for the central
executive (limited capacity as it is part of the STM store). It has a limited
capacity.
Phonologic Slave system is a temp storage system + holds sound info + has a function
al loop to learn languages. 1st part- primary acoustic store: holds auditory memories
 for a few seconds then they trace decay ‘inner ear’. 2nd part- Articulatory
loop- rehearses auditory memories ‘inner voice’. Processes acoustic info-
limited capacity + duration.

Long Term Memory

3 types of LTM

1. Declarative Memory (explicit)- person can explicitly state that

they know something, Knowing that
- Mainly conscious LT memories
- Episodic memory- personal memories of events + emotional
events
- Semantic memory- shared memories of facts + knowledge
about the world
2. Non-declarative (implicit) memory- person can implicitly do
something
- Knowing how mainly unconscious l-t memories
- Procedural memory- automatic memory for a procedure/
how to do something

Declarative memory: Episodic + semantic memory

 Episodic memory- store info abt individual person. ‘Episodes’ of a

person’s life, autobiographical memories, strength of emotional links
to the strength of memory made. Stored in hippocampus
 Semantic memory- stores info abt meanings of lang, symbols +
concepts unrelated to individual. Memories are general or fact
based, stronger than episodic memories. In temporal lobe
 Tricky to separate out these 2 types

Procedural Memory

 Stored in motor cortex + cerebellum

Episodic Semantic Procedural
Explicit or Explicit/ Declarative Explicit/ Declarative Implicit/ Non-
Implicit Declarative
Type Personal experiences Knowledge of the Skills and tasks
world
Brain Hippocampus and Temporal lobe Motor cortex and
Region prefrontal cortex cerebellum

Bower et al (1969) Exp

 Tested p’s ability to freely recall 4 lists of 28 words, p’s had to write
as many as they could remember using free recall (any order)
  1 group of p’s were given the words categories + averaged 73/112
words
 2nd group had the words presented in a random order + recalled
21/112 words on avg.
 Semantic LTM where we store words by meaning

Evidence to support types of LTM

 Prefrontal lobes, mainly in left hemisphere are responsible for

episodic memories
 Case studies have shown that some p’s have problems with
episodic but not semantic memory
 Patient HM- intact LTM but impaired STM in diff brain regions
 He could implicitly get procedurally better at exp task- supports
that procedural + declarative memories are in diff regions of
brain- distinct stores
Brain scanning techniques – fMRI, MRI
and PET scans
Allows psychologists to view diff parts
of the brain that are active during diff
memory recall tasks
Episodic memory – hippocampus and
temporal lobe
Semantic memory – temporal lobe
Procedural memory – cerebellum (fine
and gross motor skills and balance)

Alzheimer’s patients

◦ It has been shown that these patients can sometimes form new
episodic memories but cannot form new semantic memories.
◦ Therefore, this supports that although these two memories are both
declarative memories, they must reside in different parts of the
brain
◦ Also, these two types of long-term memories are affected differently
in Alzheimer's and that the stigma that ‘all memories are lost’ in
Alzheimer’s may not be completely accurate.
◦ Belleville (2006) found that patients with mild cognitive
impairment who received training performed better of episodic
memory tasks than those who received no training.
◦ Give’s economic implications for the rehabilitation of memory
loss
Strengths Weaknesses
Scientific and accurate- Case Study data- unique, brain damaged
brain scanning data patients + data may not always be
An understanding of generalised, lack of control over variables
the types of LTM is May be something else that affects our
useful for memoires- too reductionist + linear. Simplistic
psychologists, exp to a complex mental process
therapists + patients w Often artificial tasks- no ecological validity
memory issues
Bower used
quantitative data- easy
to analyse

Theories of Forgetting

Forgetting Inability to recall something you have previously learned.

Memories in LTM memory are not always retrieved
3 reasons:
Due to competing memories, 2 lots of
 Proactive Interference
info become confused in memory
 Retroactive Interference
 Retrieval failure due to absence Where memories cannot be retrieved
of cues because there are no sufficient cues to
access them
Proactive Interference

 Old info stored in LTM interferes with learning of new memories,

happens when new memories are similar to memories already
stored
 New info is similar to old info (e.g. new phone number)

Key Study Keppel and Underwood (1962), (like Peterson & Peterson)

- P’s presented with meaningless three letter trigrams (i.e. GSF)

- Trigrams were presented at different length intervals
- To prevent rehearsal, P’s had to count backwards
- Results: P’s remembered the trigrams that were presented first
therefore demonstrating proactive interference.
- Conclusion: Older info interferes with learning of newer info

Retroactive Inference

◦ New info interferes with the recall of old memories in LTM

◦ For example, once new number is learnt, it’s difficult to recall your
OG number

Evidence for Interference Theory (Baddeley and Hitch)

◦ Retroactive interference tested with rugby players, recall all teams

they played against that season, due to injury some missed games
 ◦ Hypothesis- those who played the most games would forget more
than the others who didn’t play all games
◦ Found to be true, supports the retroactive interference. The newer
team names played interfered with the recall of the older team
names played.
◦ This can also be true because the time frame for the season was
short and so interference is more likely when the time for memories
is shorter.
Strengths Weaknesses
There is supporting evidence from Not ecologically valid research- lab research
experiments such as trigrams don’t represent everyday
Supporting evidence from exps: proactive- memory
Keppel & Underwood, retroactive- Baddeley Simplistic exp of forgetting- only explains
& Hitch how similar info/memories are forgotten
Practical applications e.g. in revision- avoid Individual differences play a part in
learning similar topics/subjects together forgetting too

Explanations for Forgetting: Retrieval Failure due to Absence of Cues

 Memories cannot be retrieved bc there are not sufficient cues to

access them
 Failure to adequately retrieve memories from LTM could be a reason
for not remembering all from LTM, Need ‘cues’ to retrieve memories
from LTM
 Cues can be any trigger from the environment (senses), 2 types:
context-dependent forgetting, state dependent forgetting
- Acc to Tulving (1974) Info abt the physical surroundings (external
context) + abt physical/psy state of learner (internal context) is
encoded when we learn info
 Encoding specificity principle- recall of memories is most
efficient in same location, context + coding when memory was
learned,
 This is because internal + external cues are stored at the same time
for LTM
 Coding is the way info is transformed into a format that can be
stored and retrieved from memory

Context- Can occur when the environment during recall is diff

dependent from environment you were in when you were
forgetting learning
State- dependent Mood or psy state during recall is diff from mood you
forgetting were in when learning

Godden and Baddeley (1975)

 ◦ Context-dep memory- improved recall of specific episodes or info
when contextual cues relating to environment are same during
encoding + retrieval.
◦ Importance of setting for retrieval, 18 deep-sea divers had to
memories 36 unrelated words of 2/3 syllables
◦ Words in same setting were better due to conditions and external
context acted as a cue
Goodwin et al

 State-dep forgetting- Mental state of learning can affect recall of

that memory
 48 male med students asked to remember a list of words when
drunk or sober, p’s asked to recall after 24 hours when some were
sober but others had to get drunk again
 Info learnt while drunk is more available when in the same state
later, same state- better memory
Evaluation

Controlled exp- can be replicated- reliable Limited eco validity- artificial

Real-life applications- strategy to improve tasks
recall in eye-witness testimony- asked to Demand c’s- knew they were in
describe model a study
Simplistic- no cause and effect

Eyewitness Testimony:

 Misleading information- Leading questions – Loftus and Palmer

(1974), Post-event Discussion – Gabbert (2003)
 Anxiety – Loftus (1979) and Johnson & Scott (1976

Leading Questions- What is eyewitness testimony?

◦ Aka eyewitness memory.

◦ Includes encoding of the memory to LTM, may be partial- crimes are
fast, stressful and can induce anxiety- affects attention
◦ Witness retains info in STM + rehearses info over a period of time,
Most info will be forgotten (due to 7+- 2)
◦ Memory retrieval which may happen with help of retrieval cues or
questioning
Loftus and Palmer Study

◦ Post-event info which will have an impact on the storage + retrieval

of their memories of crime and reduce reliability of testimony they
give
◦ Aim- explore how info provided after an event, in the form of leading
q’s may affect memory, 2 separate lab exps
Experiment 1

- 45 students from Uni of Washington shown 7 vids of car crashes in a

random order, from 4-30 seconds- made for driver education
courses (researchers aware of car speeds)
- P’s given a standardised questionnaire asking them to give an
account of film they just saw, and then asked q’s (mainly filler + not
analysed by researchers)
◦ Critical q ‘How fast were the cars going when they__________
into each other’
◦ P’s split into 5 conditions- only change was the verb used which
indicated a diff level of crash, Smashed, Collided, Bumped, Hit,
Contacted
◦ Loftus + Palmer argue that memories are constructed based on 2
types of info: what we perceived as happening at the time, info
received after. And that post event info can distort memories
Conclusions

◦ 1. Distortion of memory has occurred because it has been

reconstructed: the witnesses schemas may have affected their
memories
◦ 2. Response-Bias: The participants may not have been sure about
the speed of the car to start for, and therefore may adjust their
answer to fit with the expectations of the researcher (this is a form
of demand characteristic).
◦ 3. Leading questions affected the accurate recall of the event
Experiment 2

◦ In the first experiment, there was uncertainty over whether leading

questions used had led to a reconstruction of the participants’
memories, or whether they had simply led them to give certain
answers due to demand characteristics.

Second Experiment

◦ Researchers aimed to show more conclusively that post event info

can distort memories
◦ 150 students shown a 1 min film of a multiple car crash, 3
conditions- how fast were the cars going when they smashed/hit
into each other? + no q on car speed
◦ Participants were recalled one week later and asked another series
of questions about the film, mostly filler or distraction questions
◦ Critical q in second questionnaire: Did you see any broken glass?
 ◦ P’s in condition where verb ‘smashed’ was used led them to falsely
remember broken glass (as ‘smashed’ implies that glass was
broken), lowest speed estimation when word ‘contacted’ was used
Overall Conclusion

1. Reconstructive memory hypothesis is supported - info gathered at

the time of an event is modified by data gathered afterwards.
2. Car crash eyewitnesses are not very good at estimating speeds of
vehicles involved
3. Leading questions can influence the memory of eyewitnesses.
Summary

◦ Memories are fragile and easily: distorted, affected by schemas and

by post-discussion info + misleading info

Evaluation

Highly useful study – impact in the courtroom, Lab experiments – may lack ecological
eyewitness testimony can no longer be used validity
as the only evidence to convict someone Watching videos of car crashes is not
Highly scientific study – validity + good use of the same emotion/stress/anxiety as a
controls real life crime so low ecological validity

Post Event Discussion

◦ Memory of an event/crime may be contaminated by convos the

witness has
◦ Crime or accident is discussed after, misleading info in the real
world can come from co-witnesses
◦ Includes police questioning and effects of repeated interviewing.
Where words or phrases of interviewer might become part of the
memory, esp if interviewer is not skilled at questioning and uses
leading q’s
◦ Conformity effect- where co-witnesses confer their stories of the
crime + come to a unanimous story/consensus, memories- distorted
+ inaccurate
Gabbert (2003)

◦ Aim: to investigate the effects of post-event discussion on

eyewitness testimony
◦ Procedure: 60 university students and 60 members of the local
public community watched a video of a girl stealing money from a
wallet
◦ P’s were tested individually (control group) or in pairs(experimental
group)
 ◦ Pairs could discuss crime together and deceived into thinking they
watched same vid but acc diff, All P’s completed a questionnaire
about the crime
◦ Results: 71% of the witnesses put into pairs recalled false
information about the crime. 60% identified a girl as guilty –
despite the fact they hadn’t witnessed this girl commit the crime!
Conclusion

◦ Post event discussion has a powerful and negative effect on

accuracy of EWT
◦ Reason for this distortion of memory could be because of
assimilation of new info overriding memory of event + leads to
forgetting of older memory of crime (retroactive interference)
◦ Or it could be because of pressure from other witnesses/social
groups (NSI) and believing that other witnesses are right/ more
believable (ISI).
◦ So police in real life must not let witnesses discuss crimes with each
other
Eyewitness Testimony- Anxiety

Biological Approach: The fight and flight stress response

 A psychological change in response to a real or imagined stress,

hypothalamus in the brain sends a signal to the adrenal glands to
release adrenaline (prepares body to fight and flight- dilate pupils,
increase heart rate and blood pressure)
 Fight + flight stress response can impact cognitive functions
including memory
Johnson and Scott (1976)

 Aimed to investigate the effects of anxiety on EWT

 Independent measures design, lab experiment, p’s waiting in a
reception for the ‘experiment’ to start
◦ Group 1 (control/no weapon) – P’s overheard a conversation about
faulty lab equipment and witnessed a man running past holding a
greasy pen
◦ Group 2 (experimental group/weapon) – P’s overhead a heated
exchange, heard glass smash and witnessed a man running past
holding a bloodied knife
◦ All shown 50 photos + had to identify the man running past, told
that the man may or may not be in the photos, accurate id- control
49% + 33% exp group
◦ Those exposed to a weapon/knife had higher levels of anxiety so
focused their attention on the knife and not the face of the man
 ◦ Loftus (1979)– later called this the ‘Weapon’s focus effect’, anxiety
of seeing a weapon reduces EWT accuracy

Evaluation

Lab experiment and highly controlled, but mimicked Demand characteristics- p’s there
a real-life crime as events in reception area was the for an exp so perhaps expecting
experiment something to happen
IMD so less demand characteristics than if RMD P’s deceived- ethical guidelines
Highly controlled support- when a weapon is used, broken- protection of harm of p’s
it induces anxiety, witness focuses on weapon not + effect of seeing a weapon
on criminal appearance- EWT reliability lowered

Contradictory Study- Yuille and Cutshall (1986)

◦ Investigated the effect of anxiety in a real-life shooting incident

where one person was killed, and another was seriously injured.
◦ 21 original witnesses of the crime + 13 witnesses agreed to take
part (aged 15-32), interviewed 4-5 months later, still recalled same
detail + OG testimony
◦ Furthermore – the witnesses did not succumb to bias of leading
questions and the anxiety they felt at the time of the shooting did
not seem to impact their memory recall of the event,
◦ Therefore, refuting the weapons focus effect and demonstrating
that witnesses respond differently in real life compared to
experiments.
◦ Christianson (1993) supports this and found victims of genuine bank
robberies were more accurate in their recall than bystanders.
Further Evaluation of EWT
 Early research support from Bartlett (1932)- memory is not a preserved snapshot
of event like a vid but reconstructed
 Affected by our cognitive schemas so we may recall items/events/people due to
our schema/exps + not a real memory
 Memory is a confabulation- more like stories, with details added/subtracted based
on level of confabulation
 ‘War of the ghosts’- each time story was read, it got shorter + details were lost
 Supports that there are factors that affect the recall of EWT

Discussion: What is the Yerkes-Dodson effect?

The Yerkes-Dodson Law (1908) proposes that:
 An increase in psy arousal improves
performance on any given task but only up to a
point
 ◦ Once arousal has passed a critical level called the optimum point,
performance tends to decline.
◦ We sometimes say we are ‘too stressed’ to perform well.
To conclude

◦ Meta-analysis review of studies looking at EWT + anxiety,

Deffenbacher 1983 found that ½ stated that high anxiety-> good
recall and other ½ low anxiety-> good recall
◦ So Yerkes- Dodson effect could explain that a moderate level of
anxiety is optimum for recalled eye witnessed events
◦ Moderate level of anxiety- good for exam recall, should we induce a
moderate level of anxiety to EWTs when they recall a crime (Cue
dependent forgetting)
Catastrophe Theory

◦ The theory proposes that, in some cases, alongside high

physiological arousal, feelings of anxiety become so
intense that a sudden, catastrophic drop off in
performance occurs.
General conclusion on effect of anxiety on EWT
◦ Results from studies are inconclusive
◦ The Yerkes-Dodson law seems to explain how some people recall
better with an optimal level of anxiety
◦ Anxiety can have both a positive and negative affect on the recall of
EWT
The Cognitive Interview
Improving accuracy of cognitive interview
Developed by Geiselman and Fisher in 1985 and in to criticism that the
‘standard police interview’ was too police led and could lead to inaccurate
EWT
Geiselman identified 4 key principles for the CI:
1. Cognitive reinstatement
2. Report everything
3. Recall from different perspective
4. Recall in reverse order
Building rapport, time + trust with witness/victim is important for
interview is to work.
Cognitive Reinstatement Recall Everything
Mentally recalling the context of the event (i.e. Witness encouraged to recall every
weather, smells, sounds, feelings, other people). detail no matter how trivial
Linked to state-dependant and context dependent
remembering
Cues in the environment are important encoding All detail encouraged- could waste
specificity principle, equally cues in mental state time or reveal important info
are important too
Recall from different perspective, another Witness recalls the event in
person's perspectives i.e. from the offender's reverse order
perspective
Can aid use of diff recall cues, found to help Avoids liars as they tend to trip up,
accuracy could help recall and prompt cues

Enhanced Cognitive Interview:

◦ Uses open q’s
◦ Builds rapport + trust
◦ Ask neutral questions
◦ Start with broad questions then prompt for specifics (funnel
approach)
◦ 80/20 rule (only 20% of talk is the interviewer)
◦ Use of imagery techniques (different view, different order)
◦ The interviewer reviews the interview (in the words of the witness)
◦ The interview is closed (like a debrief) and witness is left feeling able
to contact interviewer easily
Research Support for CI: Geiselman (1985)
◦ Aim: to investigate the effectiveness of the C.I
◦ Method: N=89 students watched a video of a violent crime and two
days later were interviewed using either the standard interview or
cognitive interview
◦ Results: the C.I yielded significantly increased correct information
than the standard interview, num of incorrect errors was the same
for both groups
◦ Conclusion: the C.I is more effective at improving the quantity of
info recalled
◦ The C.I does not increase the amount of incorrect information
recalled
Lots of other research to support the use CI is still susceptible to misleading info
of the C.I with eyewitnesses Fisher (1989) Centofanti & Reece (2006) study, p’s given
– compared CI to SI. 16 experienced post event discussion + false info following
a video of a bank robbery
detectives recorded interviews from real
P’s recalled the misleading info during C.I
life crimes. Kohnken (1999)
2 groups- 8 continued to use SI but 8 were Meta analysis of 42 studies looking at the CI
trained in the CI, interviews recorded + (approx 2500 interviews)
analysed. More recall was found in the CI but some of
Trained detectives yielded 46% more info it was incorrect
than the detectives using the SI.
90% + info recalled from the witnesses
was accurate!

CI and Age- Holliday (2012)

 ◦ Looked at effects of CI in young adults and older adults
◦ Recall of short staged crime from a film then some misinfo given to
the P’s
◦ A few days later asked to recall in a CI
◦ Asked to recall in a CI a few days later, young adults recalled 3x
more correct info
◦ Older adults not affected by misinformation, Misinformation
effect
Issues:
- Requires special training- time consuming + can use police time
- Many police forces don’t go beyond rudimentary basic C.I training-
costs + time
- Economic implications- time + cost hinder effective use of CI
- Culture bias, or gender bias in the use of the CI. Personal
characteristics of the interviewer could impact the accurate recall of
the event
- The CI has only been tested in western cultures