What you'll learn

• What Loftus and Palmer (1974) did in their classic study on eyewitness memory.
• How leading questions can affect speed estimates and later recall.
• How to evaluate the study using methodology, ethics, real-world application and contradictory evidence.
• How to apply the study to exam scenarios and research-methods questions.

The cognitive starting point: memory is active

The cognitive approach studies internal mental processes such as memory, perception, attention, thinking and language. These processes cannot be observed directly, so psychologists infer them from behaviour.

A key idea for this study is that memory is not like a video recording. Instead, memory can be reconstructed: when you remember an event, you may combine the original information with later information, expectations and schemas.

A schema is a mental framework built from past experience. For example, your schema for a car crash may include broken glass, loud noises, high speed and injury. If a question strongly suggests a severe crash, that schema may affect what you report.

Why this study matters

The named classic cognitive study is Loftus, E. and Palmer, J. C. (1974), “Reconstruction of automobile destruction: an example of the interaction between language and memory.”

It is especially important because it investigates eyewitness testimony: evidence given by people who have witnessed an event, often a crime or accident.

For example, “How fast was the car going when it smashed into the other car?” suggests a more violent crash than “How fast was the car going when it contacted the other car?”

AO1: Loftus and Palmer (1974)

Aim

Loftus and Palmer aimed to investigate whether the wording of a question could affect participants’ memory of a car accident.

More specifically, they wanted to see whether different verbs, such as smashed, hit or contacted, would change:

• estimated speed of the cars
• later memory for details that were not actually present, such as broken glass

Experiment 1: the verb and speed estimates

Sample

The sample was 45 university students.

This is important for evaluation because students are a specific group, so findings may not generalise to all eyewitnesses.

Procedure

Participants watched seven short film clips of car accidents. After each clip, they answered questions about what they had seen.

The critical question was:

“About how fast were the cars going when they ______ each other?”

Different groups received different verbs:

• smashed
• collided
• bumped
• hit
• contacted

Variables

In Experiment 1:

Findings

The more intense the verb, the higher the average speed estimate.

Conclusion

The wording of a question affected participants’ judgements. “Smashed” produced the highest speed estimate, while “contacted” produced the lowest.

However, Experiment 1 alone does not prove memory was changed. It could be that participants remembered the same accident but gave a higher estimate because the verb encouraged them to respond differently.

Experiment 2: did the question alter memory?

Experiment 2 was designed to test whether leading questions could change later memory, not just immediate estimates.

Sample

The sample was 150 students.

Procedure

Participants watched a film of a car accident. There was no broken glass in the film.

They were divided into three groups:

One week later, all participants were asked:

“Did you see any broken glass?”

Findings

Participants in the smashed condition were more likely to report seeing broken glass, even though there was none.

Conclusion

Experiment 2 supports the idea that misleading post-event information can alter memory. The word “smashed” may have activated a more severe crash schema, making participants more likely to “remember” broken glass.

Response bias or memory alteration?

Loftus and Palmer discussed two possible explanations.

Response bias

Response bias means the wording influences how participants answer, without necessarily changing their memory.

For example, “smashed” may simply make participants think the researcher expects a higher speed estimate.

Memory alteration

Memory alteration means the wording changes the memory trace itself. Later information becomes integrated into the original memory.

Experiment 2 is stronger evidence for memory alteration because participants later reported a detail that was not present.

AO2: applying the study to scenarios

In an exam, you may be given a scenario involving a police interview, a witness statement or a teacher asking a student what happened. Your job is to link the scenario to Loftus and Palmer.

Research methods links

Loftus and Palmer is useful for Component 2-style thinking because it is an experiment with clear variables and quantitative data.

Levels of measurement

In Experiment 1, estimated speed is interval-level data. In Experiment 2, “yes” or “no” to broken glass is nominal data.

Inferential test choices

For Eduqas, you should choose inferential tests using the aim, design and level of measurement:

For Experiment 2, if you were comparing the frequency of “yes” and “no” responses across groups, chi-square would be appropriate because the data are nominal.

For example, with chi-square, Spearman’s rho and t-tests, the observed value usually needs to be equal to or greater than the critical value. For Mann-Whitney U, Wilcoxon and the sign test, the observed value often needs to be equal to or less than the critical value.

A one-tailed test is used when the hypothesis predicts a specific direction. A two-tailed test is used when the hypothesis predicts a difference or relationship but not the direction.

A Type I error is a false positive: concluding there is a significant effect when there is not. A Type II error is a false negative: failing to find a significant effect when there really is one.

AO3: evaluating Loftus and Palmer

Strengths

High control

The study was carried out in a controlled setting. Participants watched the same film clips and answered standardised questions. This improves reliability because the procedure can be repeated.

It also helps establish cause and effect: the verb was manipulated, and changes in speed estimates or false memories were measured.

Quantitative data

The study produced numerical data, such as mean speed estimates and counts of “yes” responses. This makes comparisons clear and objective.

Real-world application

The findings have important applications to police interviews and court cases. They suggest interviewers should avoid leading questions and use neutral prompts such as “Tell me what happened.”

This links to the development of better interviewing practices, including encouraging free recall and reducing interviewer contamination.

Weaknesses

Low ecological validity

Watching a film clip is not the same as witnessing a real accident. Real eyewitnesses may experience fear, surprise, injury or personal involvement.

This means the findings may not fully apply to real crimes or accidents.

Sample bias

The participants were students. They may be younger, more educated and more used to psychological tasks than the general population.

This reduces population validity.

Demand characteristics

Participants may have guessed that the wording was important. If so, they may have changed their answers because they thought that was what the researcher wanted.

Contradictory evidence

Yuille and Cutshall (1986) studied witnesses to a real armed robbery. They found that eyewitness accounts were often accurate months later and were less affected by misleading questions.

This suggests Loftus and Palmer’s findings may apply more strongly to artificial laboratory situations than to all real-life eyewitness memories.

Ethics

Loftus and Palmer used film clips of car accidents, so there was some possible risk of mild distress. Under the BPS Code of Ethics and Conduct, key issues include:

• Consent: participants should know broadly what they are taking part in.
• Deception: full aims may not have been revealed, otherwise participants might guess the hypothesis.
• Right to withdraw: participants should be able to leave the study.
• Protection from harm: accident films should not cause excessive distress.
• Confidentiality: participants’ data should be kept anonymous or confidential.
• Debrief: participants should be told the true aim afterwards, including that there was no broken glass.

Overall, the ethical risks were relatively low, but deception and possible distress still need to be considered.