By Ruthanna Gordon
Watch the latest CSI, and you’ll see a world where evidence is clear-cut and scientific. By the end of any episode, the viewer can feel secure that the guilty party has been caught, and will be tried and convicted.
In the real world, the situation is often far more complicated. If investigators are very lucky, criminals leave behind sufficient DNA for testing, and either have that DNA already on file from previous convictions, or get caught so that a new comparison can be made. (Although unless the case is very high-priority, test results come back in weeks rather than hours.) Under these circumstances, determinations of guilt can be clear-cut and scientific. Other types of evidence tend to be fuzzier.
Eyewitness testimony, for example, is far more unreliable than people like to admit. Leading questions can change someone’s memory outright, or bias their descriptions. One study showed people a video of a car accident:
How fast do you think the cars were going when they smashed into each other?
If you’re like most people watching the video, your answer to that question—presumably asked by the prosecution—will be 10-15 miles per hour faster than if the defense asked you about the speed of the cars when they “contacted” each other. Other studies have used similar manipulations to change memory for weapon used, appearance of the perpetrator, and sometimes even the perpetrator’s recollection of their own actions.
The same biases can also affect those whose job it is to interpret evidence. The whorls and lines of your fingertips really are unique—but fingerprints tend to be smudged. Fingerprint experts learn towards finding matches between the prints collected at a crime scene, and those they already have access to. And their confidence levels for these judgments tend to be unrealistically high.
An understanding of these issues has begun to trickle into the courtroom. Defense lawyers sometimes bring in memory researchers to warn juries about eyewitness fallibility. Juries are also more likely to receive information about accuracy rates for tests, or how well one can really see a face at 200 feet in the fog. And yet, it’s still common for people convicted on other evidence to be exonerated by DNA testing. It’s not yet clear how to create a working justice system that responds appropriately to these findings—but we’ve certainly got a long way to go.