I know a bit about surveys, methodology, and how to draw conclusions from statistics, and the union's conclusion extrapolates that just because 75% of voluntary respondents had been or had witnessed bullying, that percentage can be deemed for the whole workforce, is just plain wrong.
I’ve refrained from putting my pennies worth (until now..!) in on this subject for the very reason you state above – this survey is worthless by any sensible measure and tells the readers nothing, it has no value whatsoever other than to cause CC to now look for and
expect bullying where there is none.