Hacker Newsnew | past | comments | ask | show | jobs | submitlogin

Yes, I do think it's less than 0.0036%.

First we see from the data that CPS declared 2.5 million of those 3.2 as "non-victims" so we can eliminate type 1 errors from almost 80% of the population right there.

Of the remaining 700k it's not inconceivable at all that in a big percentage of cases there were actually significant parental problems. Frankly I bet that type 2 errors are much more common that type 1 for this population.



http://www.theatlantic.com/national/archive/2014/07/in-a-yea...

> 2.5 million of those kids were declared 'non-victims.' Another 686,000 were 'abused' or 'neglected.' And an estimated 1,640 kids died as a result.

Their error rate on non-victims is a helluva alot higher than that.

I don't have faith their error rate for victims is much better given we don't have data for that.




Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact

Search: