Reliability of

cognitive
processes
Syllabus

Reconstructive memory
• Human memory is not an exact copy of events but rather a reconstruction that may be
altered over time, through discussions with others or input from the media. Research
shows that memory may be changed during storage, processing and retrieval, for
example, due to schema processing. Knowing this could be important not only in our daily
lives but in particular in relation to eyewitness testimony.
• Factors related to the study of reconstructive memory
• Relevant concepts related to studying the nature of reconstructive memory could be but
are not limited to:
• confabulation—a memory based on a fabricated, distorted or misinterpreted memory
often believed to be true in spite of contradictory evidence
• schema processing—memory processing based on prior knowledge in the form of
schemas which could result in distortion
• • false memories—recalling an event that never happened and believing it to be true.
2
General Outline

3
Reliability of cognitive processes:
reconstructive memory
Unreliability of memory
• limitations to reliability of memory: retrieval of info from LTM may depend on whether
or not you are using a particular schema
• retrieval/ non-retrieval of info depends schemas in use is only one dimensional
unreliability of memory
• another dimension: tendency of memory to be distorted
• As Bartlett (1932) showed, cultural schemas allow us to reconstruct remembered
stories and events out of our own past experiences
• although we may have shared those experiences with friends and family, it only
takes the action of asking two siblings to remember the same event from their
childhood to make us realise that we reconstruct them differently from each other
• Each of us has a unique experience of an event and a personal way of recollecting it
through our own schemas.
9/3/20XX Presentation Title 4
The theory of reconstructive memory &
eyewitness testimony
• theory: memory— x passive retrieval of info. from the long-term , but is an active
process that involves the reconstruction of info the theory of reconstructive
memory
• reconstruction literally: construct the memory again
• reconstructive nature of memory is important for our day-to-day living, but it is
even more important when it comes to questions of eyewitness testimony
• The California Innocence Project was formed in 1999, and was the first of
Innocence Projects worldwide
• One of the main issues on which their lawyers focus is that of eyewitness misidentification of
suspects
• Once eyewitness convictions began to be overturned by DNA evidence, the realisation
developed that as many as 1 in 4 stranger identifications by eyewitnesses could be wrong
and that eyewitness misidentification played a role in more than 70% of the convictions that
were later overturned by DNA testing in the USA (California Innocence Project, 2017)

5
Post-event Information

• The classic Loftus and Palmer (1974) study into eyewitness

testimony illustrates the reconstructive nature of memory
• At the end of their journal article, Loftus and Palmer state: ‘(I)t is
natural to conclude that the label, smash, causes a shift in the
memory representation of the accident in the direction of being
more similar to a representation suggested by the verbal label’
(p.588)
• similar to schema theory, and it could be argued that ‘smashed’ suggests a
‘serious accident’ schema that then triggers the higher estimate of speed,
and the suggestion of broken glass at the scene
• However, it is worth noting that Loftus and Palmer’s study itself is not an
investigation of schema theory.
9/3/20XX Presentation Title 6
RESEARCH DESCRIPTION:
Loftus and Palmer (1974)
RELEVANT TOPICS:
Eyewitness testimony
AIMS OF THE RESEARCH:
To investigate the effect of leading questions on the accuracy of eyewitness testimony
PROCEDURES:
Experiment 1
•45 students were split into 5 groups of 9 & shown film recordings of traffic accidents (shown 7 films)
•order in which the films: different for each participant
•following each film: participants were given a questionnaire asking them to answer a series of questions about the accident
•most of the questions on the questionnaire: distracters, one critical question: ask about the speed of the vehicles involved in the collision
•question varied among the 5 groups of participants
• ‘about how fast were the cars going when they hit each other’
• others: ‘hit’ was changed for ‘smashed’, ‘collided’, ‘bumped’, ‘contacted’
•accuracy of estimates: x matter
• numerous studies hosed the people are not very good at judging how fast a vehicle is travelling
• all participants watched the same films, yet they gave significantly different mean speed estimates
•response bias: subject might be uncertain whether to say 30mph or 40mph & a verb of a higher intensity ‘smashed’ biases the response to a higher estimate
• memory of the event in this case does not change
•memory change: question causes a change in the subject’s memory representation the accident e.g. verb ‘smashed’ alters the memory so that the subject remembers the
accident: more sever than it actually was

Experiment 2
•rationale: memory actually undergoes a change, subject more likely to remember other details that did not actually occur, but fit well into the newly constructed memory
•150 participants: shown a film depicting a multiple-car accident
•given a questionnaire that included distracter questions & 1 critical question
•participants were split into 3 groups: smashed into each other, hit each other & control group (c critical question)
•one week later: subjects were given a questionnaire (x film)
• consists of 10 qs, 1 critical qs — ‘Did you see any broken glass?’ (yes/no), but it had not been any broken glass

7
RESULTS/FINDINGS:
Experiment 1
•the estimated speed was affected by the verb used
•e.g. participants who were given the verb smashed reported an average speed of 40.5mph
•participants who were given the word contacted reported an average speed of 31.8 mph
•overall difference of 8.7mph

Exp 2
- did you see any broken glass?
• yes: 32% when the verb smashed was used, only 14% when the word hit was used (almost the same as the 12% in the control group)
• actually there weren’t any broken glass
• higher-intensity verb led to a higher speed estimate & higher probability of recollecting an event that had never actually occurred
EXPLORATION (EVALUATE RESEARCH):
Conclusion
Exp 1
•show that the accuracy of eyewitness testimony is affected by leading questions & a single word in a question can significantly affect the accuracy of our judgements

Exp 2
•The verb smashed has connotation of faster speeds and broken glass and this question led the participants to report seeing something that was not actually present. Their memory for the
original event was distorted by the question used one week earlier, demonstrating the power of leading questions.

•visual & verbal information might be stored separately & leading question might have interfered with the verbal storage, but not visual one
• remained to be seen whether or not verbal post-event info. could be integrated with visual info. obtained originally at the time of the event
•real-life eyewitness testimony often involves recognition (recognising a stimulus) rather than a recall (in the absence of a stimulus)
• leading qs (with verbs of varying emotional intensity): provide post-event information that contributes to reconstructive memory
• real-life: post-event info. might take more aggressive forms e.g. providing a person with misleading information

Loftus and Palmer’s research has questionable ecological validity.

•questioning participants about everyday events like a car crash appears to be a genuine measure of eyewitness testimony —> unethical to create real car crashes
•However, the participants in their research watched a video of a car crash and witnessed the events unfold from start to finish. In everyday reports of car accidents, witnesses rarely see the
whole event; they are either involved in the event directly, or see a small part of the event happen in their peripheral vision.
•Therefore, their results to do reflect everyday car accidents , unable to conclude if participants involved in real accidents, who would have a stronger emotional connection to the event,
would also be susceptible to leading questions in the same way.
•Demand characteristics: Participants corrected their original answer according to the chosen verb.
•lacks population validity.
•Their two experiments consisted of 45 and 150 American students. It is reasonable to argue that the students in their experiment were less experienced drivers, who may be less accurate at
estimating speeds.
•unable to generalise the results to other populations, for example, older and more experience drivers, who may be more accurate in their judgement of speeds and therefore not as
susceptible to leading questions.

However, Loftus and Palmer’s research took place in a laboratory (low ecological validity) of Washington University and was therefore highly controlled. This high degree of control reduces
the chance of extraneous variable, increasing the validity of the results. Furthermore, it is easy for psychologists to replicate their research, to see if the same results are achieved with a
different population.
8
RESEARCH DESCRIPTION:
Loftus, Miller & Burns (1978)
RELEVANT TOPICS:
Eyewitness testimony
AIMS OF THE RESEARCH:
To investigate how verbal info. supplied after an event influences a witness’s visual memory for that event
PROCEDURES:
•recognition procedure: 195 students from the University of Washington— shown a series of 30 colour slides depicting successive stages in a car-pedestrian accident
• slides featured a red Datsun travelling along a side street toward an intersection with either a stop sign (for half the subjects)/ yield sign (other half)
• remaining slides: Datsun turned right & knocked down a pedestrian who was crossing at the crosswalk
•after viewing the slides, subjects answered a series of 20 questions
•question 17: Did another car pass the red Datsun while it was stopped at the stop sign? (half), Did another car pass the red Datsun while it was stopped at the yield
sign?
•2-by-2 experimental design
•half the subjects received consistent post-event info. while for the other half info was inconsistent with their actual visual memories
•filler activity: a forced-choice recognition test, administered in which subjects were required to pick the slide they had actually seen from a pair of slides
RESULTS/FINDINGS:
•subjects who received misleading post-event info. able to correctly recognise the slide actually seen in 41% of the cases
•subjects who received consistent post-event information: correct choice in 75% of the cases

EXPLORATION (EVALUATE RESEARCH):

•variation of this experiment: both the questionnaire & recognition task were administered not immediately but a week later
•correct recognition of the slide in the inconsistent-information group: less likely
•misleading post-event verbal information: integrated with visual info. from before—> reconstructive memory
•more time passes from the moment of the original event, the stronger effect misleading info. has on our visual memory
•verbal post-event info. actually can integrate with visual info & alter it

9
RESEARCH DESCRIPTION:
Yuille and Cutshall (1986)
RELEVANT TOPICS:
Eyewitness testimony (real life)
AIMS OF THE RESEARCH:
To look at how accurate witness accounts and and the type of errors made within them.
PROCEDURES:
•Yuille and Cutshall interviewed real witnesses of a real crime.
•The witnesses had observed a gun shooting incident on a spring afternoon in Vancouver, Canada.
•A thief entered a gun shop, tied up the owner, and stole money and guns. The owner freed himself and picked up a revolver. He went outside to get the car’s registration
number, but the thief wasn’t yet inside and he fired 2 shots at the store owner.
•After a short pause, the owner fired 6 shots from his revolver - killing the thief. The store owner recovered from serious injury.
•The witnesses saw the scene from various locations - passing cars, buildings and from the street.
•This incident was chosen by the researchers because there were enough witnesses to compare their accounts. Plus, the money and weapons were recovered from the thief's
dead body so there was a lot of forensic evidence to verify witness testimonies. Research wouldn’t interfere with a case because the death of the thief closed the file.The
previous police questioning wouldn’t interfere with the study as there were many visible elements to the scene that the police hadn’t focused on.
•21 witnesses were interviewed by the police after the incident. 13 of these agreed to the research interview.
•The store owner (the victim) didn’t want to relive the trauma. Contact was attempted with the other 7 witnesses but 2 had moved away, and the other 5 didn’t want to take
part.
•The researchers had verbatim (word for word) reports of the police interviews. Participants in these interviews had been asked to describe the events in their own terms and
then the officer asked a series of questions to amplify what had been said. They were recorded by hand.
•The 13 participants were interviewed 4 months later. Their responses were recorded and transcribed. They gave an account and answered questions (following the same
procedure as the police interviews).
•2 misleading questions were asked: One about a/the busted headlight, the other about a/the yellow quarter panel (which was actually blue).
•They were asked about the degree of stress they experienced at the time on a 7 point scale. They were asked about their emotional state before the incident, and problems
like sleeplessness afterwards.
•The researchers used a careful scoring procedure to compare the details from the police and research interviews and what actually happened.
•The details were divided into ‘action details’ and ‘description details’.‘Description details’ were further divided into ‘object details’ and ‘people details’.
•Some difficulty with the scoring occurred.
RESULTS/FINDINGS:
•The researchers gained more details overall (especially object details) than the police did. This was because the researchers asked things which were of no interest to the
police.
•The police gained more action details and person details.
•There was variation in what the witnesses reported as they’d seen different amounts of the incident.
•7 were central witnesses and 6 were peripheral witnesses. Both groups were equally accurate. In the police interviews, the central witnesses accuracy was 84.56% while the
peripheral accuracy was 79.31%.
•Errors were still relatively rare and the accuracy remained high 5 or 5 months later.
•The misleading information had little effect on the witnesses. 10 out of 13 of them said there was no broken headlight or yellow quarter panel, or that they hadn’t noticed
those particular details.
10
EXPLORATION (EVALUATE RESEARCH):
•Yuille and Cutshall was the first investigation in witness testimony to use real witnesses of a real incident.
•Eyewitnesses aren’t as inaccurate as laboratory studies suggest.
•The long-term accuracy was possibly due to the incident being memorable and unusual.
•The researchers suggested they may be investigating flashbulb memory - where a specific and relevant event is recorded in memory in great detail.
•They remembered more as they were directly involved. Laboratory studies wouldn’t capture this involvement.
•The field nature of this study undermined the findings of laboratory ones about the effect of leading questions. Attempts to mislead didn’t succeed.
•Also, the witnesses felt more adrenaline than stress during the event and whatever stress they felt didn’t negatively affect memory.
•They showed that one detail being wrong shouldn’t mean the entire account should be rejected.

Evaluation:
Generalisability:
•There was only 13 research witnesses and only one unique event was studied.
•Generalising findings to criticise laboratory experiments may be unfair. Flashbulb memory is different to what was tested in laboratory experiments.
Reliability:
•The findings seem reliable as care was taken to make sure the testimonies never altered what actually happened.
•Limited reliability due to the small sample. There's a possibility of participant variables and sample bias.
•The researchers replicated the police interviews to a great extent and did achieve similar results. So it appears the study is replicable and as a result reliable.
Applicability:
•The study and its findings have applications within the reliability of eyewitness testimony.
•The findings can be applied to the reliability of cognitive interviews debate. The study supports the use of cognitive interviews as they include open questioning which allows
more details to be drawn from witnesses. The narrative styled research study gathered more information than the previous police interrogations.
•The findings can also be applied to flashbulb memory theory. The participants in this study recalled the details of the incident better than in other studies because they had
actually witnessed it.
Validity:
•High validity as it’s a field study (real environment and real situation) - it's true to life.
•The scoring turned qualitative data into quantitative data. This incurs a chance of bias with subjective interpretation which decreases the validity of the results.
•The comparison between the police and researcher interviews was controlled by the rigorous scoring system.
•High ecological validity as it investigated real eyewitness testimony.
Ethics:
•Consent - The 13 research witnesses all consented to being a part of the research.
•Deception - There was attempted deception with the misleading questions, however the deception was minor and was mostly unsuccessful.
•Right to withdraw - 5 of the potential 20 research participants exercised their right to not be involved in the study.
•Confidentiality - It can be assumed the identities of the participants were kept confidential.
•Protection of participants - The participants weren’t harmed in the study. The store owner was not forced to participate in case of reliving traumatic memories.
•Brief/Debrief - They were not given a full brief because there was deception during the procedure. It is unclear whether the witnesses were later debriefed.

11