Wolfgang's citation for "paternity tests" in which there is a dispute - or potential one - about paternity is a good one.

The numbers (>>10%) that I gave come from a vastly different "population," though.

I haven't seen specific numbers. For people innocently attempting to show links to ancestors via DNA testing, higher percentages have been reported - but only in "generic" terms. The numbers reported for these tests quite probably include broken links within a few generations back, as well as simple father/child "errors," and I haven't seen anyone publishing reviewable analyses of the results. The results would, of course, report an "error" of this kind for nearly all adopted children, many of whom may never have been told about the adoption(?).

It is all largely hearsay, so I'd suggest caution in quoting any specific percentages as confirmed "facts."

Any "errors" found in this genealogical testing certainly may say as much about the (lack of) reliability of genealogical records (and especially about "family histories") as about the behaviour of "wayward wives."

John