2.2.1 Reconstructive Memory

↳

Key Definitions

Misleading question: one of the forms that post-event information can take; suggest information that is not entirely consistent with what actually happened

Post-event information: information about an event provided (directly or indirectly) after the event already occurred

Recall: a form of retrieval, retrieval of required information from memory in the absence of any prompts

Recognition: a form of retrieval that involves identifying an object as previously seen

Reconstructive Memory: the theory that views memory to be an active process of recreation of past events as opposed to a passive process of retrieval

Essential Understanding

• Theory of Reconstructive Memory

→can explain certain phenomena like cases of memory unreliability: including memory distortions; when you remember things that didn’t happen (or not exactly what the way they happened)

→ suggests that memory is not a passive retrieval of information from a long-term store, but an active recreation of the event in the mind every time it is remembered

→ recognizes two kinds of information:

↳ information obtained during the perception of the event

↳ external post-event information

→ over time, these two information sources can get integrated to the extent that we are unable to tell apart

• Support for Reconstructive Memory

→can be seen in classic study done by Loftus and Palmer (1974): showed that in an eyewitness situation people’s accounts of an event can be influenced by slight differences in the way the question was made

↳discussed two potential explanations:

↳ genuine memory or simple change bias (when the memory is unchanged, but participants tweak their responses based on what they think is expected of them)

→ phenomenon or reconstructive memory was found both for recall and recognition tasks, both for verbal and visual information

Ex: Loftus, Miller, and Burns (1978): involved a recognition task that found that misleading verbal post-event information can integrate with visual memory and distort it

→ Criticism: lack of ecological validity

→ Yuille and Cutshall (1986): conducted a study in naturalistic settings (a real-life store robbery) and found that misleading questions had very little effect on how eyewitnesses recalled the events, contradicting reconstructive memory theory

↳ results may still be a reflection of a completely different memory mechanism→ flashbulb memory

Researches

• Loftus and Palmer (1974)→ the eyewitness study

↳key: in an eyewitness situation, misleading post-event information can integrate with memory of the event and alter it

Aim: investigate if memory can be altered by misleading post-event information (in an eyewitness situation)

Method: experiment; independent measures design

Participants: university students, convenience samples

Experiment 1: 45 students to 5 groups

Experiment 2: 150 students to 3 groups

Procedure:

Experiment 1:

↳ Participants were shown recordings of traffic accidents and then were given a questionnaire with a number of questions about the accident they just witnessed

→ Only one critical question: “About how fast were the cars going when they hit each other?”

→ Five groups in experiment 1 only differed in the emotional intensity of the verb used in the sentence

↳ in other groups, “hit” was changed to “smashed”, “collided”, “bumped”, and “contacted”

↳ IV was therefore the misleading post-event information, operationalized as the emotional intensity of the verb in the question

↳ DV was the speed estimate

Experiment 2:

↳ Participants were shown a film of a car accident and then were then asked to fill out a questionnaire

→ Three groups of participants got three different versions of the critical question:

↳ “smashed into each other”

↳ “hit each other”

↳ no critical question (control group)

→ A week later, participants were given another questionnaire with one critical question with a “yes/no question”: “Did you see any broken glass?”

→ IV was the emotional intensity of the verb in the leading question

→ DV was whether or not the participants reported having seen broken glass (in reality, there was no broken glass

Results:

Experiment 1:

Experiment 2:

→ Emotional intensity of the verb in the leading question influenced the probability that participants would report seeing broken glass

Conclusion:

→ Experiment 1 showed that misleading post-event information influences eyewitness accounts of an event

↳However, two potential explanations for this finding

→ There could be a change in genuine memory change that causes a change in the participant’s representation of the event

→ There could be response bias: memory doesn’t change, but verbs of higher emotional intensity causes participants to give higher estimates when they are uncertain

→ Experiment 2 ruled out the second explanation of Experiment 2

↳ verbs of higher emotional intensity may cause participants to recall events that didn’t occur

↳ researchers concluded that we should reject response bias explanation and accept genuine memory change

• Loftus, Miller, and Burns (1978)→ reconstructive memory in a visual recognition task

↳key: effects of misleading verbal post-event information can also be seen in visual recognition tasks

Aim: investigate whether verbal post-event information can be integrated with visual information obtained originally

Method: experiment, independent measures, 2 × 2 experimental design

Participants: 195 university students

Procedure:

↳ participants were shown a series of slides depicting a car (red Datsun) that approaching an intersection with a sign, then turned right and knocked down a pedestrian

→ critical slide was the one showing the sign at the intersection

↳ for half the participants, the slide shows a stop sign

↳ the other half had yield sign

→ After seeing the slides, participants were asked a series of questions, with the critical one as follows

↳ half of the participants were asked: “Did another car pass the red Datsun while it was stopped at the stop sign?”

↳ other half were asked: “Did another car pass the red Datsun while it was stopped at the yield sign?”

→ Results in a 2×2 experimental design

→ Participants had a forced-recognition test where they had to pick the slide they had seen from a pair of slides they had seen s from a pair of slides

Results:

→ Participants who received misleading post-event information (Groups 2 and 3) recognized the slide they had seen correctly 41% of the time

→ Participants who received consistent post-event information were able to do so 75% of the time

Conclusion:

→ Misleading post-event verbal information can get integrated with visual memory, alter it, and affect performance on a visual recognition task

↳ has a great practical significance: visual recognition tasks are a common practice in police investigations

• Yullie and Cutshall (1986)→gun store robbery study

↳key: phenomenon of reconstructive memory is not found in high naturalistic settings involving strong emotional reactions while witnessing an event

↳ However, in such cases we could be dealing with a separate memory mechanism that overrides reconstructive memory (flashbulb memory)

Aim: investigate whether eyewitness accounts get distorted as a function of misleading post-event information in a naturalistic setting

Method: interviews with elements of an experiment

Participants: 13 eyewitnesses to real crime (a gun store robbery) in Vancouver

Procedure:

→ In this real-life robbery, a thief entered a gun store, tied up the owner, stole money and guns, and left

→ Owner untied himself, took the gun, and ran outside

→ Then, a shooting occurred and the robber was killed

↳ this was witnessed by 21 people with different viewpoints and they were all interviewed by police

→ 4 months after the incident, researchers conducted interviews where they also used experimental elements (some had leading questions with misleading information, some not)

↳ to determine the accuracy of participants’ memories, they were compared to official police records

Results:

→ Misleading questions had very little effect on recall

→ Participants were able to accurately recall a large number of details

Conclusion:

↳ Two potential explanations:

→ Reconstructive memory is a phenomenon that is only found in artificial conditions of laboratory experiments

→ Study actually explains another memory phenomenon—flashbulb memory

↳ Flashbulb memory: occurs when the witnessed event is accompanied by a strong emotional experience, so memory of the event gets “imprinted” with high accuracy

Extra Study

→ McCloskey and Zaragoza (1985)→doubted the results of the Loftus, Miller, and Burns (1978) study

↳ they claimed that the results were likely due to response bias

→ Used the same-slide recognition procedure, but participants were shown a series of 79 colored slides

↳ slides showed a maintenance worker entering an office, fixing a chair, finding 20 dollars, stealing it, and a calculator, and leaving the office

→ Series of slides had 4 critical slides with 3 versions

Ex: maintenance worker either used a hammer, wrench, or screwdriver

Afterwards, participants read a text with misleading information (i.e. some of the texts mentioned a “tool” when the participants saw a hammer

→ people in the control read a narrative mentioning wrench when they saw a screwdriver

Another key difference was the forced choice of activity the participants saw

Ex:

↳ some participants would see a hammer in the slides, but read about a wrench in the narrative, but also had to choose between a hammer and wrench

Results and Conclusion:

→ In control group, participants correctly recognized the “tool” 69% of the time, a higher % than expected by chance if we assume participants don’t remember→ so participants were guessing

→ In the second group, they recognized the correct object 40% of the time, a lower % than expected by chance→ misleading information was integrated with prior visual info

→ In the third group, participants correctly identified the object 66% of the time→ nearly no difference to the control group

Key: concluded that presentation of misleading information didn’t effect the original memory