Just One Minute
Balanced Fare: We Report, You Deride

Thursday, September 26, 2002

ABC News on the Central Park Jogger

This story has it all. My only quibble: in discussing Reyes's decison to come forward at this time, they do not seem to mention that the statute of limitations has now passed for this crime, so his confession has no consequences for Reyes.

ABC News on False Confessions

I presume that the experts are not as foolish as they appear in this interview, so I will blame it on the interviewer. But honestly, the basic logical fallacy of this presentation is laughable. Briefly, the ABC text presents the experts as saying: "we have studied false confessions; certain interrogation techniques are used in false confessions; we see these techniques used here; therefore, these confessions may be false."

Please. Try this example: "All men are mortal"; "I have identified something that is mortal"; "therefore, I have identified a man". Well, actually, I have identified one of my kid's goldfish, and I need to take care of that.

As to our experts, they need to study valid confessions - crimes where the criminal confessed, but external evidence corroborates their guilt - compare these to "false" confessions, and identify the differences. If the themes that are common to false confessions also run through true confessions, then it is back to the drawing boards.

An example: How might we get a false confession? "One method is by asking leading questions...", we learn. And do police sometimes, often, always, or never ask leading questions in a situation that ultimately leads to a true confession? No relevant information is provided to address this point, but my guess, based purely on crime shows, is that they do.

As I said, not having checked the underlying research, I blame ABC News for this one. Max Power had a related point about innumeracy and false confessions.

UPDATE: OK, here is a bit of research from one of the experts shown on ABC. These are extensive footnotes to a paper I can't find online. Soundbites: The author recognizes the "true confession" / "false confession" problem I mention. Serious methodological problems with implementing such an approach. Final work based on a study of 200 "false" confessions. That's not good. And, he explains why it is impossible to estimate a "false confession" rate.

Comments: Post a Comment