Monday, June 10, 2013

CPAG - messy associations

Now here's an interesting clash.

Today the Child Poverty Action Group (CPAG) released a new report about the link between poverty and child abuse (more to say about the content soon).

The report talks more than  once, and negatively, about an Auckland University Study which found 83 percent of children with a substantiated finding of abuse by age five had appeared in the benefit system by age two.

Perhaps more alarmingly, Volume 2 of the While Paper contains a chapter on “identifying children in target populations” (Chapter 4). Despite the difficulties outlined in this review in identifying when and where child maltreatment might happen, this chapter introduces the idea of a ‘predictive risk model’ to assist in identifying ‘vulnerable’ children (Vaithianathan et al., 2012). The model includes 132 variables for inclusion in the core algorithm (Vaithianathan et al., 2012, p. 11), making it arguably of limited use. The researchers used data based on benefit receipt, and found a strong association with benefit receipt and child maltreatment, although it is unclear if the beneficiary was identified as the abuser. This use of data to target subgroups of the population raises very serious questions about marking out and branding families on the basis of factors or circumstances over which they have  no  control.  It  suggests  child  maltreatment  is  a  function  of  membership  of  particular  social groups,  something  for  which  the  evidence  is  very  weak. 13  

Named as one of the project team behind the Auckland University study is Claire Dale, Research Fellow, Department of Economics, Auckland University. She is one of the et al cited above.

Well, that's a surprise. Claire Dale is also a member of CPAG, and Virginia Dale was the CPAG 2012 Student Scholar (supervised by Claire Dale) who produced this highly political piece, Myths and Facts: Sole Parents and the DPB. (See my response to that paper here).


Psycho Milt said...

It suggests child maltreatment is a function of membership of particular social groups, something for which the evidence is very weak.

That's astonishingly incompetent for supposedly educated people.

1. Actually, it suggests child maltreatment is correlated with membership of this group, not a function of it.

2. Trying to dismiss a strong piece of evidence for something by saying the evidence for it is very weak just makes a person look stupid.

JC said...

"Perhaps more alarmingly, Volume 2 of the While Paper contains a chapter on “identifying children in target populations”

Of course its alarming to CPAG. Identifying problem groups and dealing with the problem makes CPAG redundant.