Let's assume: you're a witness to a car accident. Two vehicles collide at an intersection. A police officer approaches and asks: “At what speed did the cars smash into each other?” You answer—say, 60 km/h. Now imagine the question was different: “At what speed did the cars touch each other?” You might say 40 km/h. Even though you saw the exact same thing.

This isn’t a coincidence. It’s the effect of question wording on memory. And the most famous study that proved it is called the Car Crash Experiment.

What exactly happened in the experiment?

In 1974, psychologist Elizabeth Loftus and her colleague John Palmer showed a group of students a short video: two cars colliding at an intersection. Then participants were asked a single critical question—but with different verbs.

• One group: “At what speed did the cars smashed into each other?”
• Another: “At what speed did the cars collided?”
• A third: “...bumped?”
• A fourth: “...hit?”
• A fifth: “...contacted?”

The results were striking:

Word in the question / Average estimated speed (km/h)

• smashed: 65.5 km/h
• collided: 63.2 km/h
• bumped: 61.1 km/h
• hit: 54.7 km/h
• contacted: 50.8 km/h

The stronger the word, the higher the speed estimate. Even though the video was identical.

But that’s not all.

A week later, participants were asked a new question: “Did you see any broken glass?”

In reality—there was no broken glass in the film. However:

• From those who heard “smashed” — 32% said they saw glass.
• From those who heard “hit” — only 14% said they saw glass.

One word—and a person invents a detail that never existed.

Why does this happen?

This phenomenon is called the misinformation effect. The brain doesn’t record memories like a video camera. It reconstructs them every time we recall something.

When you hear the word “smashed,” your brain automatically adds associated images: a heavy impact, shattered glass, a loud crash. And these images weave into the original memory, as if you actually saw them.

Memory isn’t an archive. It’s a story you rewrite every time.

Is this only about car crashes?

No. Loftus conducted dozens of studies. For example:

• People were convinced they got lost in a mall as a child — and 25% started “remembering” details that never happened.
• Others were told they swallowed a wasp in lemonade — and began avoiding certain drinks.

It even works in therapy. If a therapist suggests, “Maybe you were abused as a child?” — a person might start “remembering” violence that never occurred. Such cases of "recovered memories" destroyed families in the 80s and 90s.

What about eyewitnesses in court?

This is where it gets scariest. Loftus showed: if a witness is asked “Did you see THE gun?” (with the definite article “the”), they’re more likely to say “yes”—even if there was no gun. But if asked “Did you see A gun?” — there are fewer errors.

Because of her research, interrogation rules for witnesses changed in the U.S. It is now prohibited to ask questions that lead toward a specific answer.

What does this mean for you?

• Your memory isn’t proof. You can sincerely believe you saw something—and be wrong.
• Words shape reality. How you ask questions—to yourself or others—changes memories.
• Double-check yourself. Ask: “What did I actually see? What did I add later?”

Conclusion

The car crash experiment isn’t just a curious fact. It’s a reminder: we don’t see the world as it is. We see it as we’ve been taught to see it.

And one word can change everything.

Sources

• 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.
• Loftus, E. F. (2005). Planting misinformation in the human mind: A 30-year investigation of the malleability of memory. Learning & Memory.