Can data quality affect R,Rfree difference?

[Gerard 'CD' Kleywegt]

i know from experience that it can unfortunately ;-)

if the resolution is the same for both datasets, how
about completeness and perhaps i/sigma ?
e.g., if the completeness is much lower you will
have fewer reflections, poorer observable to
parameter ratio, probably more overfitting and
therefore a larger difference between r and rfree

how well do the two datasets merge ?

--gerard