Experiments

Loftus and Palmer misleading questions 1974

People asked how fast cars 'smashed' into each other gave higher speed estimates than those asked about cars that 'contacted', and later more often reported broken glass that was not there.

MemoryExperiment

Elizabeth F. Loftus, John C. Palmer

Main paper: Loftus, E. F., & Palmer, J. C. (1974). Reconstruction of automobile destruction: An example of the interaction between language and memory. Journal of Verbal Learning and Verbal Behavior, 13(5), 585–589. https://doi.org/10.1016/s0022-5371(74)80011-3

The question

Can the wording of a question change what people remember about an event?

Design: experiment

The researchers set the conditions and compare outcomes between or within groups. Each profile says whether participants were randomly assigned.

Between-subjects comparison of question wording. This profile does not state how participants were assigned or give counts.

Two comparisons across two experiments

Experiment 1: contacted

  1. Watch accident clips
  2. Estimate speed using a weaker verb

Result reportedLower speed estimates

Experiment 1: smashed

  1. Watch accident clips
  2. Estimate speed using a stronger verb

Result reportedHigher speed estimates

Experiment 2: smashed

  1. Watch the film
  2. Receive the stronger speed question
  3. Later answer a broken-glass question

Result reportedMore reports of nonexistent glass

Experiment 2: comparison

  1. Receive the hit wording or no speed question
  2. Later answer the same glass question

Result reportedFewer reports of glass

Original diagram drawn for this site from the descriptions in the sources. It is a simplified summary, not a figure from the paper.

Procedure

In the first experiment, students watched film clips of traffic accidents and then answered questions about them. One question asked how fast the cars were going when they hit, with the verb changed across groups, for example 'smashed' or 'contacted'.

In a second experiment, groups saw a film and were asked about speed with one verb or no speed question. A week later they were asked whether they had seen broken glass. There was none in the film.

Findings

Speed estimates changed with the verb in the question: the stronger verb 'smashed' produced higher estimates than a weaker verb such as 'contacted'. In the second experiment, the stronger wording was linked to more reports of broken glass that had not been in the film. The textbook section on memory describes both results.

Interpretation

The authors proposed that information in the question can be combined with the original memory, so later reports reflect both. This became a starting point for research on the misinformation effect.

Limits on what it shows

  • Participants watched staged film clips, not real events, so stress and personal stakes were absent.
  • A different speed estimate might reflect a guess shaped by the wording, not a changed memory. Telling these apart requires further tests.
  • Student participants and short delays limit how far the results extend to courtroom witnesses.

Ethics

Participants viewed film clips of accidents, with little risk. The finding carries ethical weight in another way: it shows how an interviewer's wording can unintentionally shape a witness's report.

Replication and reassessment

The two experiments address different outcomes: speed estimates and later reports of glass. A wording effect in a report does not alone distinguish a changed memory from a response strategy. Keep the manipulation, delay and outcome separate when comparing later studies.

Think it through. An original practice question written for this site.

A police interviewer wants to avoid shaping a witness's memory. What kind of question could they ask instead of a leading question about a red light?

Show a way to reason about it

An open question such as 'What did you see at the intersection?' lets the witness supply details. A leading question about a red light assumes a fact, much like a loaded verb does. This is practice reasoning, not interview guidance from a source.

Nothing you think or select here is recorded.

Textbook connection

Covered in this sectionOpenStax Psychology 2e, 8.3 Problems with Memory

Section describes the car crash study.

Sources and access scope

Each label says how much of the source was opened while preparing this profile. An unopened source is listed so the claim can be checked, not because it was read.

  1. Loftus, E. F., & Palmer, J. C. (1974). Reconstruction of automobile destruction: An example of the interaction between language and memory. Journal of Verbal Learning and Verbal Behavior, 13(5), 585–589. https://doi.org/10.1016/s0022-5371(74)80011-3Consulted sourceOriginal researchNot inspected; citation only

    DOI for the original article, which has not been inspected for this profile. The verb and broken-glass findings rest on the textbook section.

  2. Spielman, R. M., Jenkins, W. J., & Lovett, M. D. (2020). Psychology 2e. OpenStax. https://openstax.org/books/psychology-2e/pages/8-3-problems-with-memoryTextbook contextSelected textbook passages inspected

    Textbook context. The section describes the car crash study, the verb manipulation and the broken glass question.