Patterico's Pontifications


A Summary of My Complaints Against the L.A. Times’s Handling of DNA, Cold Cases, and Statistics

Filed under: — Patterico @ 1:47 pm

The following is an excerpt from my 2008 L.A. Times Year in Review:


All year, the paper’s editors have been engaged in a holy war against the use of DNA in criminal cases. It started in May, when the newspaper ran an article about statistical probability in cold hit DNA cases, and it was immediately clear that some of the assertions didn’t make sense.

The editors don’t seem to like DNA when it’s used to convict.

For one thing, the article seemed to assert that larger databases made cold hits less reliable, when it would seem that the opposite would be true — at least in cases where the search revealed only one hit. A statistics professor named David Kaye agreed with me on that point. In addition, he told me, the article had falsely portrayed an anti-prosecution view of the statistical question as the consensus view — when, in fact, there is a competing view more favored by peer-reviewed articles. (The author of the L.A. Times article wrote me to claim that he had acknowledged there is a lack of unanimity of opinion, but the article didn’t clearly express this.)

But the biggest error was a flat-out statistical misstatement in the article. Professor Eugene Volokh outlined the problem. I drafted a letter to the article’s authors, and ultimately sent this e-mail about the misstatement. Then I noticed yet another error in the article, again having less to do with the math, and more to do with how the math was expressed in English. Of the three errors I identified, the paper corrected only a trivial arithmetical error, leaving the more significant misstatements standing.

The editors denied they’d made a misstatement, even though they admitted that it would be wrong to make a different statement that my readers overwhelmingly agreed was identical.

Although editors denied that they had described the statistics incorrectly, they did start describing them correctly — which I took as a silent concession that I was right.

But true vindication came when a statistics expert — one whom the paper had previously quoted as an expert — claimed in a scholarly article that the paper had “mischaracterized” the statistic that I had complained about. I once again wrote the Readers’ Representative, citing the expert’s opinion. She didn’t give me the courtesy of a reply.

A second DNA kerfuffle began when the paper ran a front-page story portraying certain matches in an Arizona database as shocking. Why, the paper suggested, the results defied the laws of statistics! Only on the back pages were readers told that most of the matches “were to be expected statistically.” One of the authors of “Freakonomics” later pronounced himself surprised that the matches were largely to be expected; apparently, like many readers, he had been misled by the article’s initial spin.

A local jury freed a clearly guilty man accused of rape; the foreman was heard expressing concerns about the case based on “recent controversies” about DNA — a clear reference to the L.A. Times‘s misleading series of articles.

In discussing a technique called familial searching, the paper did its usual shtick with DNA: it played up phantom privacy concerns, and buried the fact that the technique has been used to free wrongly convicted individuals.

One Response to “A Summary of My Complaints Against the L.A. Times’s Handling of DNA, Cold Cases, and Statistics”

  1. […] know that the L.A. Times’s series of flawed articles on DNA, cold cases, and statistics is a longstanding complaint of mine. In May 2008, The Times ran an article about the case of John Puckett that claimed that “the […]

    Patterico's Pontifications » DNA Expert: L.A. Times “Mischaracterized” Key Statistic in Front-Page DNA Article (e4ab32)

Powered by WordPress.

Page loaded in: 0.1428 secs.