Eyewitness testimony: misleading info

• Loftus and Palmer (1974) Leading questions

  Procedure: 45 participants watched clips of car accidents then answered questions about speed. Critical question: 'About how fast were the cars going when they hit each other?'. Five groups of participants, all given different verb: hit, contacted, bumped, collided or smashed.
• Findings: Verb 'contacted' produced mean estimated speed of 31.8mph. For smashed, mean estimated speed was 40.5mph.

  The leading question biased eyewitness recall of an event. The verb 'smashed' suggested a faster speed of the car than 'contacted'.
• Why do leading q's affect EWT?

  -Response bias explanation: Wording of question has no enduring effect on the memory but influences kind of answer given.

  -Substitution explanation: Wording of question does affect memory, interferes with original memory, distorting accuracy.
• Gabbert et al. (2003) Post-event discussion.

  Procedure: Paired participants watched video of same crime, each saw different angles, seeing elements their partner could not. Discussed what they saw before individually completing recall test.
• Findings: 71% of participants wrongly recalled aspects they did not see but had heard in discussion.

  Control group - No discussion, no subsequent errors. Evidence of memory conformity.
• Why does post- event information affect EWT?

  -Memory contamination: Mix information from own memory with information from other witnesses.

  -Memory conformity: Go along with others to win social approval or because they believe other witnesses correct.
• ++Real world applications in criminal justice system. Consequences of inaccurate EWT serious. Loftus (1975) argues police officers should be careful in phrasing questions. Psychologists sometimes expert witnesses in trials and explain limitations of EWT. Therefore psychologists can improve how legal system works and protect from wrongful convictions based on unreliable EWT.
• --Evidence challenging the substitution explanation. Sutherland and Hayne (2001) found participants recalled central details better than peripheral ones, even with misleading questions. May be because attention was focused on central features, so resistant to misleading info. This is not predicted by the substitution explanation.
• --Evidence doesn't support memory conformity. Skagerberg and Wright's (2008) participants discussed film clips they had seen. Participants recalled a blend of what they had seen and what they had heard. Suggests that the memory itself is distorted through contamination by post event discussion and is not the result of memory conformity.
• --Demand characteristics. Lab studies give researchers high control over variables (high internal validity), but lab experiments suffer from demand characteristics, participants try to help by guessing answers they don't know, (low internal validity).