Clever Hans, police lineups and crime labs

Clever Hans was a German horse in the early 1900’s who drew large crowds for his seemingly uncanny ability to do math (including square roots), tell time, identify musical intervals and name people. He communicated by tapping his hoof. Eventually it was discovered that he couldn’t perform these feats unless he could see the questioner and – most significantly – unless the questioner knew the answers. Hans was picking up subtle, unintentional cues from the questioner. This is a remarkable story, considering how difficult we’ve always found it to get horses to respond to even our most explicit, intentional cues, such as get off my foot.

The “Clever Hans” effect is why scientific experiments are conducted “blind.” Knowing the “right” or desired answer can unconsciously bias the testing. It’s been applied to police lineups, which are demonstrably less likely to create a misidentification when conducted by a cop who doesn’t know which one is the suspect. Although, needless to say, the NYPD lineup procedures haven’t changed since the days of George Raft, other more forward-looking police departments have adopted blind lineups.

Strangely, we’ve never noticed crime laboratories using blind procedures. The lab technician complacently testifies that she tested a sample from “the rape victim” or the drugs “recovered from the defendant.” The technician has been told the “facts” of the case, with no concern that she’s taking cues, subtle or otherwise, from the prosecution.

An article called “Getting Forensics Right” by Radley Balko proposes a radical solution to the conviction-oriented bias of state crime laboratories: set up a system of “rivalrous redundancy,” whereby every few pieces of testable evidence in criminal cases would be sent to a private lab in addition to the state lab. The state lab wouldn’t know which of their tests were being reviewed. This would keep the state lab on its toes.

Putting some competition into the criminal justice system is a good idea. Monopoly guarantees mediocrity. If only we could somehow apply it to the courts.

Do ye ken junk science?