Loftus + Palmer 1974

• Aim?

  • to determine whether information received after the event, in the form of leading questions, would be integrated into a persons memory
• hypothesis?

  • subsequently presented information would form part of the memory and cause the event to be recalled in a consistent way with subsequent information given
  • the stronger the verb used in describing the crash, the higher the speed estimate
• experiment 1 method?

  • lab experiment
  • independent groups
  • IV: verb used
  • DV: speed estimate
• experiment 1 participants?

  • 45 pp’s
  • split into 5 groups of 9
  • (hard to generalise from a group of 9)
• experiment 1 procedure?

  • pp’s watched 7 film clips showing a traffic accident taken from Evergreen Safety Council/Police department
  • clips were taken between 5 and 30 seconds long
  • film clips were counterbalanced
  • pp’s received a questionnaire after each clip: give any account of accident, answer specific questions, received critical question related to vehicle speed
• what did the questionnaire include?

  • leading questions (desired answer)
  • critical question (used to measure DV)
  • distracted question (masks focus on critical question to conceal demand characteristics)
• what was the critical question?

  • how fast were they going when they…
  • hit
  • smashed
  • bumped
  • collided
  • contracted
• results of car speed estimates

  • speed of cars never went above 40 but people estimated much higher
  • 20 - 31.7
  • 30 - 36.2
  • 40 - 39.7
  • 40 - 36.1
  • (loftus+palmer didn’t care about this)
• accuracy of speed estimate with verb?

  • phrasing of question (verb), considerably varied estimate
  • Contracted = 31.8
  • Hit = 34
  • bumped = 38.1
  • Collided = 39.3
  • Smashed = 40.8
  • (Need to know order but not number)
  • Highly significant results as there is only a 0.5% margin of error between them (p<0.005)
• conclusions?

  • differences in speed estimates could be due to;
  • response bias - verb biases the response
  • change in memory - schema accommodated to match the previous understanding of the verb
  • prediction - if the latter, it may be predicted that subjects will recall other details which did not occur but would fit with the false memory
  • construct validity -this study part doesn’t measure what they wanted to measure
• experiment 2 method?

  • lab experiment
  • independent groups
  • IV: verb used
  • DV: response to critical question
• experiment 2 method?

  • 150 students
  • 5 groups of 30
• experiment 2 procedure?

  • watched 1 clip showing a multiple car traffic accident
  • less than 1 minute video, crash was 4 seconds
  • asked the same questions again
• questions asked?

  • 50 asked how fast they were going when they hit
  • 50 asked how fast they were going when they smashed
  • 50 asked no leading question (control)
  • asked to come back a week later to do a questionnaire - “did you see any broken glass?”
• results?


  • (there wasn’t actually any glass)

  
• conclusions?

  • the questions asked subsequent to an event can cause a reconstruction of one’s memory in the event
  • pp’s took information from the original scene and then merged this information given after the event, producing a false memory
  • we are inclined to make our memories make sense and so adjust details
• discussion?

  • Loftus+Palmer proposed 2 kinds of information go into our memory

  1. info gleaned during perception of the event
  2. external info supplied after the event

  • overtime these 2 integrate - the verb smashed causes a shift in memory
• reconstruction?

  • some pp’s have forgotten whether there was glass so fill the gaps by making inferences from presented information
  • accommodative distortions
  • misremembering
  • Loftus said that memory is very fatigue
• method?

  • controlled lab experiments that fulfill scientific criteria
• data?

  • quantitative data means that the results are easily summarised and easy to compare
• ethics?

  • ethical
  • gave consent
  • knew it was a test of memory but not that they had different questions (prevents demand characteristics)
  • used staged crash without gruesome images to prevent harm/upset
• validity?

  • high design validity (controlled lab experiment)
  • knew it was a memory test (demand characteristics)
  • lab experiments lack ecological validity
  • spontaneous (memories may be different when we aren’t warned)
• reliability?

  • highly controlled lab experiment = reliable+replicable
  • both experiments provided evidence that memory of an event can be distorted by info introduced after
  • supports theory that language distorts memory
• ethnocentrism?

  • could be argued not ethnocentric as it investigates species specific behaviour
  • however, the sample does have ethnocentric bias as most students came from upper and middle social classes
• scientific?

  • controlled lab experiments are good for evidence and replication
• Usefulness?

  • shows it is possible to distort memories of eye witness to events
  • evidence has led to research into the best way for police to question witness
• nature?

  • memory is an internal process that we are born with
• nurture?

  • suggests memory can be influenced by environmental factors
• reductionism?

  • focuses on a single variable (verb used in question)
  • reduces memory recall to the cognitive processing of language and leading questions
• holism?

  • a more holistic approach could incorporate emotions, prior knowledge and social context of event
• determinism?

  • suggests that people‘s memories can be shaped or manipulated by environment
  • memory may be determined by cognitive processing triggered by external influence
• freewill?

  • there is freewill on how memory is retrieved or reported
• sample?

  • sampling bias as only students were used which causes low generalisability