Unreliability of Any Line Up???
From this week's Economist:
THE difference between fuzzy perceptions and hard science can mean everything to someone who has been falsely accused of a crime.
According to the Innocence Project, a non-profit legal clinic attached to Yeshiva University in New York, DNA evidence has so far helped to exonerate 194 people in the United States who were wrongly convicted, and who spent an average of 12 years in prison before being released.
Of the wrongly accused, over 75% had been mistakenly identified by eyewitnesses.
Since using witnesses to weed out suspects remains a crucial tool for investigating crimes, groups across the country--from crime commissions to state legislators--have been looking for better ways to run police line-ups. That effort is spotlighting an academic debate about psychology and research methods. And it has now triggered a lawsuit over a pilot study of three police departments in Illinois.
In traditional police line-ups (as featured in many a celluloid crime
drama) investigators show eyewitnesses a group of subjects or a palette of photos and ask whether the criminal they saw is among them. Many psychologists and criminal defence lawyers have called for a switch to another method that is different in two important ways. First, they say, the officers handling the line-up should not know which members of the group are "fillers": ie, innocents who have been added merely to make the eyewitness choose. That way, the witness does not pick up any cues, intentional or otherwise, about which subject the police suspect.
Second, say psychologists, eyewitnesses would respond more accurately if they could see subjects one at a time, rather than seeing a whole group at once and making relative comparisons. A number of research studies in academic settings have suggested that this method, called "sequential double blind", would lead to fewer witness errors than traditional line-ups. New Jersey's attorney-general recommended the new method to the state's police department in 2001. Since then, attorneys-general or crime commissions in Wisconsin, North Carolina and Virginia have endorsed the reforms in different ways; and legislatures in five other states are considering them.
This reform drive was called into question last year, however, by a pilot study in Illinois. Unlike earlier research, which often used video clips or other methods to simulate crimes and witnesses, the Illinois study was the first and only one to compare the two methods in the field using actual cases. During 700 or so real police investigations in Chicago, Evanston and Joliet, eyewitnesses using sequential line-ups were less likely to name the police's suspect, or anyone at all, as the criminal.
That is an ambiguous result, since the police, of course, could have been wrong about some of their suspects. But eyewitnesses using sequential line-ups named fillers as criminals--that is, they wrongly accused people who were definitely innocent--three times more often than witnesses who looked at traditional line-ups.
Many reforms remain convinced that the newer method is better, but concede that the Illinois study has slowed their political momentum. So earlier this month two reforming groups--the National Association of Criminal Defence Lawyers (NACDL), and the MacArthur Justice Centre at Northwestern University--filed a legal demand for more details about the pilot study. Scott Ehlers, who deals with state legislatures for the NACDL, says that some of its results are "fishy". The Chicago police, however, say that they are innocent. No eyewitnesses have come forward.
See this article with graphics and related items at http://www.economist.com/world/na/displaystory.cfm?story_id=E1_RSDSQPD
THE difference between fuzzy perceptions and hard science can mean everything to someone who has been falsely accused of a crime.
According to the Innocence Project, a non-profit legal clinic attached to Yeshiva University in New York, DNA evidence has so far helped to exonerate 194 people in the United States who were wrongly convicted, and who spent an average of 12 years in prison before being released.
Of the wrongly accused, over 75% had been mistakenly identified by eyewitnesses.
Since using witnesses to weed out suspects remains a crucial tool for investigating crimes, groups across the country--from crime commissions to state legislators--have been looking for better ways to run police line-ups. That effort is spotlighting an academic debate about psychology and research methods. And it has now triggered a lawsuit over a pilot study of three police departments in Illinois.
In traditional police line-ups (as featured in many a celluloid crime
drama) investigators show eyewitnesses a group of subjects or a palette of photos and ask whether the criminal they saw is among them. Many psychologists and criminal defence lawyers have called for a switch to another method that is different in two important ways. First, they say, the officers handling the line-up should not know which members of the group are "fillers": ie, innocents who have been added merely to make the eyewitness choose. That way, the witness does not pick up any cues, intentional or otherwise, about which subject the police suspect.
Second, say psychologists, eyewitnesses would respond more accurately if they could see subjects one at a time, rather than seeing a whole group at once and making relative comparisons. A number of research studies in academic settings have suggested that this method, called "sequential double blind", would lead to fewer witness errors than traditional line-ups. New Jersey's attorney-general recommended the new method to the state's police department in 2001. Since then, attorneys-general or crime commissions in Wisconsin, North Carolina and Virginia have endorsed the reforms in different ways; and legislatures in five other states are considering them.
This reform drive was called into question last year, however, by a pilot study in Illinois. Unlike earlier research, which often used video clips or other methods to simulate crimes and witnesses, the Illinois study was the first and only one to compare the two methods in the field using actual cases. During 700 or so real police investigations in Chicago, Evanston and Joliet, eyewitnesses using sequential line-ups were less likely to name the police's suspect, or anyone at all, as the criminal.
That is an ambiguous result, since the police, of course, could have been wrong about some of their suspects. But eyewitnesses using sequential line-ups named fillers as criminals--that is, they wrongly accused people who were definitely innocent--three times more often than witnesses who looked at traditional line-ups.
Many reforms remain convinced that the newer method is better, but concede that the Illinois study has slowed their political momentum. So earlier this month two reforming groups--the National Association of Criminal Defence Lawyers (NACDL), and the MacArthur Justice Centre at Northwestern University--filed a legal demand for more details about the pilot study. Scott Ehlers, who deals with state legislatures for the NACDL, says that some of its results are "fishy". The Chicago police, however, say that they are innocent. No eyewitnesses have come forward.
See this article with graphics and related items at http://www.economist.com/world/na/displaystory.cfm?story_id=E1_RSDSQPD
posted by John M. Kaman at 4:06 PM
0 Comments:
Post a Comment
<< Home