Improving the accuracy of eyewitness testimony: The cognitive interview

Geiselman et al (1984) - interviewing technique - cognitive interview (CI), based on proven psych principles concerning effective memory recall:
 - mental reinstatement of original context
 - report everything
 - change order
 - change perspective

Interview encourages interviewee to mentally recreate physical + psychological environment of original accident.

Eg may say: "Think about that day....what had you been doing...try and get a picture of it in your mind" etc - Dando + Milne (2009).

Aim to make memories accessible. People often can't access memories - need contextual + emotional cues to retrieve memories.

Interviewer incourages reporting of every detail w/o editing out, even if irrelevant. Witnesses shouldn't leave anything out, even if believe insiginificant.

Memories interconnected w/ one another - recollection of 1 item may cue other memories. Recollection of small details may be pieced together from diff witnesses -> clearer picture of event.

Interviewer may try alternative ways through timeline - changing order events occurred. Recollections influenced by schemas. If recall from end of events backwards, prevents pre-existing schema influencing what you recall.

Interviewee asked to recall incident from multiple perspectives eg by imagining how would have appeared to other witnesses present at time. Done to disrupt effect schemas have on recall. 

Suggested by Anderson + Pichert.

Meta-analysis, 53 studies, increase of 34% amount of correct info generated in CI compared w/ standard interviewing techniques - Köhnken et al (1999), although most studies involved volunteer witnesses (usually college students), tested in lab.

May be particularly useful w/ older witnesses. Neg stereotypes about older adults' memory can make them cautious about reporting info. Because CI stresses importance of reporting every detail, may overcome this difficulty. Mello + Fisher (1996) - older (mean 72 years) + younger (mean 22 years) memory of filmed simulated crime using CI or SI. CI produced more info, bug advantage of CI over SI greater for older than young.

CI designed to enhance quantity of correct recall w/o compromising quality, but Köhnken et al (1999) found 81% increase of correct info, but also 61% increase incorrect info (false positives) when enhanced CI used. Means police need to treat all info collected from CIs w/ caution - doesn't guarantee accuracy.

When used in real world, isn't just one 'procedure' - collection of related techniques. Eg, Thames Valley Police use version that doesn't include 'changing perspectives'.

Kebbel + Wagstaff report police officers suggest CI requires more time than available - prefer to use deliberate strategies that aim to limit eyewitness' report to min amount of info officers feel necessary. Also requires special training, many forced not been able to provide more than few hours - why CI not widespread.