Well, there is no magic privacy-preserving anti-maliciousness bullet -
we will use our common sense, my own years of experience in reading
survey results and reason to not be overly influenced by what looks
suspicious. But also given the volume of answers (we're at over 1800
now) and what I've seen of them, this seems quite unlikely to be doable.
If someone managed 50 answers that express the same sentiments but are
not identifiable as the same person, I will be deeply impressed ;) (and
as the survey is now closed, this is in no way a challenge!) Any way to
prevent people from doing multiple answers at the same time would also
include some sort of increased potentially privacy-infringing measures,
and I really do not want to 1. do anything that could cause participants
to be compromised 2. do anything that would make participants feel
unsafe (those two things are not the same, even if there is a lot of
overlap, but both are important).