Reconstructive Memory Model: Cognitive Psychology

The founder of reconstructive memory model

Elizabeth F. Loftus and John C. Palmer

Definition of reconstructive memory model

Memory is not 100% accurate; it retrieves fragments of an event.
Can omit details, insert false information, or change event order

Pre-Event Information

Information a person possesses or is exposed to before an event occurs.

Purpose/Effect of Pre-Event Information

Shapes expectations, attention, and interpretation of the event.

Can lead to priming or biases in how the event is perceived or remembered.

Post-Event Information

Information a person receives or encounters after the event has taken place

Purpose/Effect of Post-Event Information

Can alter memory of the original event 

Affects how the event is recalled, interpreted, and emotionally processed.

Process of recalling in reconstructive memory model

LEVELING

SHARPENING

ASSIMILATING

LEVELING

simplifying things

SHARPENING

highlighting and overemphasizing things and details

ASSIMILATING

changing details to fit our background and knowledge

The Pioneer of Reconstructive Memory Model

Bartlett 1932

• Aim: To study how cultural schemas influence reconstructive memory.

• Participants: 20 male Cambridge university students (UK).

• Procedure:Used serial reproduction – each person read and recalled the story, passing it on to the next.

• Results:Memory of the story changed in key ways:

• Omission: Unfamiliar or disturbing details (e.g., "ghosts", "black fluid") left out.

• Assimilation & Sharpening: Changed to fit cultural norms (e.g., “canoe” → “boat”, added logical connectors).

• Levelling: Story shortened (350 → 180 words).

Aim of Bartlett 1932

To study how cultural schemas influence reconstructive memory.

Participants of Bartlett 1932

20 male Cambridge university students (UK).

Procedure of Bartlett 1932

Used serial reproduction – each person read and recalled the story, passing it on to the next

Results of Bartlett 1932

Memory of the story changed in key ways:

• Omission

• Assimilation and Sharpening

• Levelling

Omission of Bartlett 1932

Unfamiliar or disturbing details (e.g., "ghosts", "black fluid") left out.

Assimilation and Sharpening of Bartlett 1932

Changed to fit cultural norms (e.g., “canoe” → “boat”, added logical connectors).

Levelling of Bartlett 1932

Story shortened (350 → 180 words).

Loftus and Palmer Experiment

Car Crash Experiment by Loftus & Palmer (1974)

Aim of Car Crash Experiment

To investigate whether leading questions can distort an eyewitness’s memory of an event (i.e., if the language used can create false memories or response biases).

Experiment 1 of Car Crash Experiment

Influence of Verbs on Speed Estimation

Procedure of Experiment 1 of Car Crash Experiment

Procedure:

• Participants: 45 American college students (opportunity sample)

• Design: Lab experiment with independent measures (5 groups).

• Task: Watched 7 video clips of traffic accidents (5–30 sec each).

• Critical question: “How fast were the cars going when they [smashed / collided / bumped / hit / contacted] each other?”

• IV: The verb used in the question.

• DV: The speed estimate (mph) given by the participant.

Findings of Experiment 1 of Car Crash Experiment

Findings:

VERB	MEAN SPEED (mph)
Smashed	40.8
Collided	39.3
Bumped	38.1
Hit	34.0
Contacted	31.8

Conclusions of Experiment 1 of Car Crash Experiment

Conclusion: 

• Wording influences memory.

•  Stronger verbs → higher speed estimates.

• Memory may be biased or altered due to language used post-event.

Experiment 2 of Car Crash Experiment

False Memory of Broken Glass

Aim of Experiment 2 of Car Crash Experiment

Aim: To test if leading questions just bias response or actually change memory

Procedures of Experiment 2 of Car Crash Experiment

Procedure:

• Participants: 150 participants, 3 groups:

1. Group 1: “How fast were the cars going when they smashed?”

2. Group 2: “How fast were the cars going when they hit?”

3. Group 3: Control – no speed question.

• 1 week later, everyone asked:  “Did you see any broken glass?” (Note: There was no glass in the video.)

Findings of Experiment 2 of Car Crash Experiment

Findings: 

• “Smashed” group more likely to say yes to seeing glass (false memory)

• Shows memory was altered, not just influenced at the moment.

Conclusions of Experiment 2 of Car Crash Experiment

Conclusion: 

RECONSTRUCTIVE MEMORY	CONFABULATION
Original memory combines with new info → distorted recall.	Falsely remembering non-existent details (e.g., broken glass).

Strengths of Car Crash Experiment

• High control of variables → cause-effect relationship established

• Replicable due to standardized procedures.

• Real-world application:

  • Informs police and legal practices.

  • Influenced the Devlin Report (1976): warned against relying solely on eyewitness testimony.

Limitations of Car Crash Experiment

Low ecological validity:

• Watching a video ≠ real-life emotional response.

• Accidents are unexpected in real life, not cued like in experiments.

Biased sample

• College students → may have limited driving experience, less representative.

Demand characteristics:

• Participants may guess the aim and try to “please” the researchers.

Confounding variable?

• Possibly just response bias (not real memory change).

Eyewitness Testimony (EWT)

When a person describes what they saw/heard during a crime.

Problem of Eyewitness Testimony (EWT)

it's reconstructive and can be changed by schemas or new info

Leading Questions

Suggestive questions that can distort memory (e.g., "the weapon" vs. "a weapon").

Impact

Can lead to false memories, wrongful convictions, or missing real criminals.