Monday, September 21, 2009

Statistics abuse in L.A.

The Los Angeles Times ran a response to NCCPR's op ed concerning the foster-care panic sweeping through the county. In claims that echo those of the county Department of Children and Family Services, almost word for word, Susan Edelstein, an adjunct assistant professor at UCLA, engages in some shameless statistics abuse.

First, she claims there is something wrong with the methodology NCCPR used to determine how many children are taken from their parents each year. That methodology is known as - counting. We measure the number of children taken away each year by looking at DCFS' own count of that very number.

That figure shows a 23 percent increase since 2003. In contrast, the adjunct professor relies on the number of children who happen to be in foster care on any given day. But that number can rise or fall for a variety of reasons unrelated to the number of children actually taken from their parents. For example, a large number of children aging out of the system with no place to go can cause this "snapshot number" as it's called, to decline.

The adjunct then compares the number of children taken from their parents in 1998 to the number a decade later and finds a decline.

We've never disputed that there was such a decline - for the first half of that decade. But the adjunct neglects to mention that the rate hit a low point in 2003 - and the current rate is, in fact, significantly higher than the rate in 2003.

The adjunct also challenges NCCPR's statements concerning a landmark study of the harm of foster care. Rather than go back and forth on this point, I hope people read it and decide for themselves.

That's why we include a link to the full study from the home page of our website, www.nccpr.org It's included in the item called "The Evidence is In."

The adjunct, who apparently does a lot of training, then says that her 40 years of experience tells her that the answer to everything is - more training. But training is no substitute for due process - especially when the trainer herself engages in statistics abuse. Perhaps it's time for the adjunct to retire and allow people with new vision and new ideas to handle the training.