Monday, September 14, 2009

Maori "identity interventions" challenged

Another study from David Fergusson et al., based on the Christchurch birth cohort of 1,265 was published in Social Policy, recently posted on-line.

It addresses the role of Maori ethnicity in childhood maltreatment. The finding that Maori children were not at greater risk of sexual abuse didn't surprise me, although the author states it is in contrast to common assumptions and other research findings. I seem to recall CYF statistics I requested under the OIA confirm Fergusson's findings.

In the area of experiencing serious and severe physical punishment and exposure to parental violence however Maori had higher rates.

Fergusson then went on to break down the Maori group into sole Maori and Maori and other. What he found here is very interesting. First is a summary of one current approach to tackling Maori over-representation in child physical maltreatment;

A third framework, focusing on the role of cultural identity has, over the preceding two decades, been the dominant explanation employed to account for the ethnic asymmetry in child maltreatment rates in New Zealand (Balzer et al. 1997, Keddell 2007, Kiro 2000, Ministry of Social Development 2006, Pihama et al. 2003, Stanley 2000, Stanley and Thompson 1999). This view proposes that it is the degree of association that Māori families have with Māori kin groupings and the level of commitment they show to traditional customary practices that will influence the likelihood of Māori children experiencing maltreatment. From this perspective, strength of Māori identity in families is a protective factor for child abuse, and a lesser identification with Māori cultural domains may increase the risk of children being exposed to maltreatment. Intervention guidelines for child abuse have therefore been specifically developed for Māori, by Māori (Kruger et al. 2004, Stanley 2000, Stanley and Thompson 1999). These focus on determining the levels of affiliation Māori families have to cultural domains and the strength of cultural identity of individuals who reside in the family. Reattachment of Māori families to cultural domains and customs has therefore become a key feature of current child abuse intervention efforts.

Then comes Fergusson's findings and the implications;

The results suggest that sole Māori identity may be a risk factor for exposure to physical child abuse and inter-parental violence....These findings do pose a challenge to current policies aimed at reducing the over-representation of Māori children in rates of child maltreatment, which emphasise “identity interventions” that are not evidence-based and are largely ideologically driven. Even though such policies are no doubt well intentioned and observe statutory requirements unique to the New Zealand context, following the view expounded by UNICEF (2003, 2007), they must be exposed to ongoing critical scrutiny and empirical evaluation.


bez said...

Ha! This makes one extremely intrigued how similar statistics would look like for crime rates, school results, DPB usage etc.
It may well mean that the standard response (increasing cultural identification, cultural awareness and cultural cohesion) is in fact counter productive. One can see that pretty high statistical significance was necessary for him to be brave enough to publish these results.

Anonymous said...

There is a major criticism of this paper when you look at the numbers. For the sole Maori (n=52) and Maori and Other analysis (N=62). Around 30% of sole Maori were found in the higher levels of physical abuse (levels 2 and 3) and around 21% of Maori and Other. This means that the finding that sole Maori are much more likely to be abusers and therefore cultural programmes aren't appropriate or should be questioned is based on a finding in few people. For sole Maori, 30% of 52 people is 15.6. If sole Maori were to have the same % as Maori and Other this would equate to 21% of 52 people which is 10.92. Thus this whole conclusion is based on a finding that there were 4.5 more people in the sole Maori category than expected at this level of physical abuse. The trend across sole Maori, Maori and Other and non-Maori was significant from a statistics point of view however as a basis to make such a strong conclusion is irresponsible use of statistics and over interpretation of the data. As a statistician, I am concerned that a positive p value (<0.05) has been used to make a far more expansive interpretation of these findings than is warranted - particularly as this has nationwide implications yet based on so few in one study. Very disappointing from such a well respected researcher and also concerning that this wasn't spotted.