Question
Do we get the right answer if we compare
irradiated and unirradiated women using
observational (ie non-randomised) data from
cancer registries?