Loftus and Palmer (1974) classic

• generalisability?

  • researchers used 45 participants from Washington Uni, 5 groups having 9 participants and they were a self-selected/volunteer sample. This sample is not representative of a target population due to being volunteers in one country of one age range.
  • second sample has a higher sample size (150) however
• reliability?

  • highly standardised, study 1 shows 7 film clips ranging from 4-30 seconds and experiment 2 shows a 1 minute film clip shown to all participants. All participants were given the same questionnaire with the same critical question
• applications?

  leading questions can influence EWT, so it must be avoided by police and lawyers when interviewing. it's useful in society since banning leading questions can prevent a miscarriage of justice
• internal validity?

  • quantitative data is collected (eg answering just yes or no for the broken glass question) so results can be statistically tested. standardised procedures also mean a decrease in extraneous variables so it's more likely the independent variable had an effect on the dependent variable (infer cause and effect)
• ecological validity?

  • artificial environment seen with an artificial video. videos aren't entirely upsetting to the population compared to an actual car crash so a lower level of physiological arousal can be seen. decrease in mundane realism means participants may not give same results in a real life environment