I spend time studying identification methods, eyewitness identification and the flaws in methods I have read a lot of Gary Wells work at Iowa State University. Police misuse of methods etc. The montage method used in the Watson investigation is the worst possible method and the easiest to manipulate to get the required identification. On that alone the trial alone would never succeed in the US, UK or Canada. It has an 80% failure rate under ideal conditions. The police in the montages used in the investigation broke every rule of best practise possible.
Photo Montages as used by local police in 1998 were under intense tests in many jurisdictions because of the huge number of false positive results they produced even if well carried out. Those done by Police to identify Scott Watson would be absolute classic examples of how not to do the montage system. If it was possible to get it more wrong it has not been found yet. Best practice for montages then was avoid then if at all possible but if they were to be use then ALL the photos used must as far as is possible match the descriptions of the alleged suspect. No one photo should stand out either by different clothing which should as far as is possible be the same. Back grounds of photos should be identical with nothing to draw attention to any one particular photo. The most important thing is the supervising officer should not know which photo is the alleged suspect. This is to prevent the officer giving subtle clues as to the suspect.
Further to the blind provided by the supervisor not knowing the alleged suspect the witness should be told that the alleged suspect may not be in the montage at all. This in some part provides a double blind. A major disadvantage of montages is that witnesses compare the photos before them and select the one that they believe is closest to the alleged suspect. In tests where witnesses were not told the suspect might not be in the montage all most every witness picked a person even if the suspect was not in the montage. When asked about how sure they were they often expressed doubt but felt they had to pick some one. Best practise is to imediately ask the witness how sure they are and record this and then remove the photos before they can look again. Time should limited for the witness to decide.
With montages if a second is required and it is preferable to have two then none of the photos in the first should be in the second montage as witnesses will focus on that even if it is the wrong person. Even in the best run montage identification trials and real use situations the false id rate was still around 70-80% in other words the wrong person was identified more times than the right one. The success rate improves if the supervisor is aware of the real suspect and even while trying not too does give subtle hints and in some case blatant hints as to the real suspect. In most police organisations photo montages are a historical embarrassment and extremely rarely used now because of the number of miscarriages of justice they have been a major factor in. Photo identification is still used but sequentially and only one photo is visible at a time and are used with the same warnings and supervisor blind and have a false Id rate of about 10-15%.
Even well organised lineups now use the same methods of double blind where the suspect may not be in the line up and the line up supervisor does not know if the suspect is there or not to prevent visual hints. The degree of certainy is recorded straight away and the witness is not given time to rethink their certainy or lack of it. Ideally all identification methods used should be video recorded as well as the witness certainty being videoed. Audio recording is acceptable as a fall back but verbal reports are not acceptable as evidence specially if the supervisor knew the suspect.
It was a surprise today to discover the New Zealand Police are still using montages as their main indentification tool in spite of their proven unreliability They have modernised their methods since 1998 but still don't follow what is considered best practise for montages double blind and video recording of the process and results and the photos used now are named although this is covered instead of lettered and an effort is made to match fillers to the alleged offenders general appearance.
Posted by lindsayrk
at 10:09 PM
Updated: Saturday, 22 November 2008 9:38 PM