Monday, August 05, 2013

Was he blinking or winking?

An article that's a couple of years old now but still pretty interesting. Taken from BPS research digest (November 2009).

When a police line-up with six one-eyed men is better than a line-up with none.

To make it fair, they gathered as much Italian plumbers,
who spoke English and looked like Mexicans.
You're mugged by a man with a patch over one eye. You describe him and his distinctive appearance to the police. They locate a one-eyed suspect and present him to you in a video line-up with five innocent "foils". If this suspect is the only person in the line-up with one eye, prior research shows you're highly likely to pick him out even if, in all other respects, he actually bears little resemblance to your mugger. So the challenge is: How to make police line-ups fairer for suspects who have an unusual distinguishing feature?

Police in the USA and UK currently use two strategies - one is to conceal the suspect's distinguishing feature (and tell the witness they've done so); the other is to use make-up, theatrical props or Photoshop to adorn the other members of the line-up with the same distinctive feature. Theodora Zarkadi and her colleagues compared both approaches and found the fairer method is to replicate the unusual feature.

Zarkadi's team presented 110 undergrads with 32 photos of real-life inmates taken from the Florida Department of Corrections website. Photoshop was used to apply distinctive features including tattoos and piercings. Six of these distinctive "suspect" offenders were then embedded, one each, in six picture line-ups alongside five previously unseen "innocent" offenders. The participants' task was to pick out the suspect in each line-up.

The key finding is that the students made significantly more correct identifications when the innocents had been given an identical distinguishing feature compared with when the suspects' unusual feature had been hidden (approx 58 per cent accuracy vs. about 39 per cent).

This advantage was replicated in a second experiment in which the suspect was sometimes absent from the line-ups (akin to what can happen in real life). In this case, when the suspect was present, identification was again more accurate when the innocents also appeared with the same distinguishing feature (approx 50 per cent vs. 30 per cent). When the suspect was missing from the line-up (i.e. six innocents appeared), the students made false identifications on about 60 per cent of occasions, but this figure wasn't affected by whether the suspect, when present, had his unusual feature hidden, or if instead his feature was replicated in the innocents.

"Police officers should be aware of this ... empirical result when constructing line-ups for suspects with distinctive features and should replicate rather than conceal these features," the researchers said.
________________________________________________________________________________________________________
Zarkadi T, Wade KA, & Stewart N (2009). Creating Fair Lineups for Suspects With Distinctive Features. Psychological science : a journal of the American Psychological Society / APS PMID: 19883492

No comments:

Post a Comment