If the results of policies are so far apart it implies that data is not correctly collected and recorded.
How can we have "Canada continues to be among the top performers in PISA 2012, with fair and inclusive policies that can contribute to high levels of equity, although performance in mathematics, reading and science has decreased across PISA cycles" if " The most recently available data
show that child poverty rates for status First Nations children living on-reserve rose to a staggering 60% in 2010. By contrast, poverty rates among Indigenous children living off reserve have improved somewhat, while non-Indigenous children have seen little change to their circumstances since 2005.
Disaggregating child poverty by identity reveals three broad groupings, or tiers, of suffering in Canada"...what kind of data is sent to OECD?