Loftus & Palmer (1974)

• Overview: To investigate the reliability of memory through seeing how information provided after an event influenced a witness’s memory of the event. The researchers changed the verb in a question when asked witnesses to recall an event.
• Participants: 45 undergraduate students from the University of Washington for exp 1, 150 for exp 2
• Experiment One (45 ppl)

  • Experiments were shown 7 clips of traffic accidents. They then had to fill in a questionnaire based on what they witnessed about the accident. The questionnaire had filler questions and one critical question, which was where they changed the verb (how fast were the cars going when they smashed/bumped/contacted each other?)
• Experiment Two (150 ppl)

  • Conducted just in case results from experiment 1 were because participants were unsure and the leading question only guided their speed estimates and hence this wouldn't be evidence that memory is reconstructive
  • Participants only watched one film of a car accident, there was only smashed and hit verbs. Participants were asked for speed estimates and whether they saw broken glass.
• Results: When the critical question had the word smashed or collided, estimates for speed were significantly higher than other words like contacted or bumped (40mph for smashed vs 31mph for contacted). So greater intensity of word, higher speed estimate.

  
  

  In the second experiment, smashed still produced higher speed estimate (10.8mph) but for both conditions, the higher the speed estimate the more likely participants recalled broken glass (which wasn't even there)
• Memories are reconstructed, as supported by the misinformation effect - informatino received after an event is later mistakenly recalled as part of the memory of that event. Because the effect was demonstrated, it shows memory is reconstructive since things aren't being accurately recorded but are actively rebuilt.