Does That Depend on What the Meaning of "Led To" Is?

"Nationwide, misidentification by witnesses led to wrongful convictions in 75 percent of the 207 instances in which prisoners have been exonerated over the last decade, according to the Innocence Project, a group in New York that investigates wrongful convictions." Exoneration Using DNA Brings Change in Legal System, NYTimes (October 1, 2007)

Does the Innocence Project therefore mean to assert that false confessions can have led to at most 25% of the 207 wrongful convictions mentioned above? Or does the meaning of the claim about the role of eyewitness identifications in the 207 convictions depend on what the meaning of "led to" is?

A (simple) math (or logic) quiz: If 75% of ALL wrongful convictions are the result of mistaken eyewitness identifications, does it follow that eyewitness identifications are mistaken 75% of the time? (Hint: The answer is "no." For example, the statistic 75% may be true and yet it may also be true that eyewitness identifications are correct 99.9999997% of the time [and it may also be true that eyewitness identifications are correct only .0000001% of the time].)

In the same NYTimes article the reporter (Solomon) reports:

“It’s become clear that eyewitnesses are fallible,” said Lt. Kenneth A. Patenaude, a police commander in Northampton, Mass., who is an expert on witness identification techniques.

Does Mr. Patenaude mean to say that until the advent of DNA, people such as judges, lawyers, law teachers, jurors, and so on believed that eyewitness identifications are infallible?

What utter nonsense!

I recall an editorial by the New York Times a few years ago that made a similar claim about confessions -- an editorial that asserted that DNA had revealed that confessions are not infallible. There too we unquestionably -- infallibly -- witnessed utter nonsense: in that instance, the ludicrous tacit assertion that before the advent of DNA people did not realize that confessions can be false. I wonder: Had the New York Times heard of Miranda, Escobedo, et al.? And in connection with the question of the fallibility or infallibility of false identifications, I wonder: Have NYTimes reporters heard of pre-DNA identification cases such as Gilbert v. California, 388 U.S. 263 (1967)? If not, perhaps NYTimes reporters -- and NYTimes editorial writers -- should get a bit of legal training.