I'd want to see the raw data, but my guess would be that
variability of intelligence is extremely sensitive to neonatal
environment. Adopted children probably have a more disruptive
neonatal environment, and even adopted children "reared
together" may have drastically non-correlated neonatal
experiences.
If the study binned data from adopted children "reared together"
who were adopted at ages as old as four or five, then that would
certainly explain the low correlation. Intelligence seems to be
far more sensitive to the environment in the first five years than
it is to environment in later childhood. A closer look at the data
and the collection methodology would be needed to resolve this
particular question.
It all depends on the age of the children at which the period
of "being reared together" began.
-- Eric Watt Forste ++ arkuat@pobox.com ++ expectation foils perception -pcd