6.4.1, 6.4.2 - Can eyewitness testimony be trusted?
Eyewitness testimony can help identify events and people, but an inaccurate account can harm both the accused and victims. This lesson uses criminological research to answer the contemporary question of when testimony deserves trust and how its collection can protect justice.
Why the question matters to society
The question is how society should use testimony when sincere witnesses can make mistakes. An incorrect identification can contribute to a wrongful conviction, while dismissing useful testimony may prevent investigation of an offence. A failure to identify the actual offender also affects victims and public safety. Psychological reasoning should therefore compare the consequences of accepting an unreliable account with the consequences of rejecting a well-supported one.
Accuracy means correspondence with the event. Reliability concerns consistency, so an account can be consistently wrong. Credibility concerns whether an account is believable and well supported. A confident manner may influence credibility judgements without establishing accuracy. Rather than asking whether every witness is trustworthy, ask what conditions produced this account and what independent evidence can corroborate it.
In a fictional case, a witness saw a person briefly at dusk, discussed the event with a friend and later received a question that assumed a red bag. A clear later description of a red bag might reflect actual perception, the friend's account or the interview wording. An investigator needs the first account, the opportunity to see and the interview history to judge these alternatives. The witness should not be accused of dishonesty merely because the account changed.
Applying memory research to the risk
Post-event information can become confused with information witnessed directly, making the history of an account relevant to its accuracy. Loftus and Palmer's controlled accident-film research showed that stronger wording increased speed estimates; a separate delayed experiment found more reports of absent glass after a smashed question than after a hit question or no speed question. This supports the concern that later reports can incorporate expectations introduced by questioning. It does not show that all testimony is false or that real-world error rates equal film-study rates. Original report.
For the fictional red-bag case, asking "What was the person carrying?" allows an answer without naming a colour or assuming a bag. Asking whether the bag was red supplies both. If the witness then repeats red across several interviews, repetition cannot establish that the original observation contained it. Recording the exact first account makes changes and possible sources visible.
Weapon focus creates a different difficulty: a threatening or unusual object may capture attention while the face is poorly encoded. Even a neutral later interview cannot recover detail that was never adequately processed. Experimental restaurant-slide evidence links weapon viewing to attention and poorer memory, while leaving threat and unusualness as competing explanations. Author-hosted weapon-focus report. This supports asking about viewing opportunity, rather than assuming that being close to an event means having attended equally to everything.
Continue this lesson.
You’re previewing NovaLearn without an account. Create a free account to read the rest.
Create free account