Maybe you misunderstood what it was about absence that was not harmful?
http://psycnet.apa.org/index.cfm?fa=...1973-04138-001
This says interactions are very affected, but not "traditional measures of sex-role typing" so maybe whatever had little harmful effect was not absence over all measures but a particular one?
I can only think it was something like that (or the point was being made that that statement is NOT true) as it is clear that your point of view is correct:
http://depletion.blogspot.com/2011/0...on-female.html