A Correction To A Correction

Or maybe it's a clarification. Anyway, in a May 23 post I wrote:

"In a recent post, I wrote:

"I had an email exchange the other day with the lead author of the schizophrenia and violence paper. He assured me that, in terms of absolute risk, schizophrenics were not more violent than the rest of the population. He said he's troubled to see how the TAC people and others are using his results. "They know better," he wrote."

There are two errors in that graf. I wrote absolute risk. I should have written relative risk. Yep, there's a difference. Two, I not only misquoted my correspondent, but I misapplied the bad quote to the wrong point. The correspondent was not criticizing TAC.

I regret the errors."

Let me gingerly tread into the world of stats and risk once again. In terms of risk (the simple form as most of us understand it):

1-year prevalence of any violence in community residents (from a 1990 paper by Jeffrey Swanson, et al.):

With no mental disorder: 2.05% were violent
With schizophrenia: 12.69% were violent

There are differences between relative risk and absolute risk--and anyone who really wants to know the precise difference should go check the Wikipedia entry on risk, which is way too complicated to summarize here. Suffice to say, that when it comes to absolute risk the large majority of people with schizophrenia do not commit acts of violence.

Hopefully, I have it right this time. BTW, if anyone thinks risk is dicey business, look into the four meanings of the term "mean" (known as "average" to most of us).