RemieV said:...she will be able to identify a box containing a male who speaks Russian...
... she will be able to identify a box ... amongst many ... identical boxes if there is a Russian male inside. None of the other boxes can contain people...
HiPavel
since
you
seem
to
have
time
to
post
how
about
updating
the
forum
on
your
progress
?
Inthe
appropriate
thread
of
course
.
Pavel, Thanks for the update - next posts should be in your own thread please.
Please remember (everyone) to keep posts in this thread specifically on the topic of the claim in question.

If pavel do can write posts like above, then that would be good both for the forum and for himself. Any idea how the trick is done?
So a series of say 15 tests (a maths expert needed here) where there are 20 boxes, with some where there is a man who cannot speak Russian, some where there is a man who can speak Russian, some where there is a woman who can/cannot speak Russian, and some where there are 20 empty boxes might settle this. If she gets the Russian speaking man right x times out of y tests, she passes.
I assume that she cannot talk to the boxes, or try to weigh them, or touch them, or interact with them in other ways, but the above may be a starting point for a protocol - rather time consuming though at one hour per test.
I'm not sure why you'd want to complicate things so much. Finding the person out of 20 possible boxes has a 1/20 chance.