This is the most revealing statistic when you compare the two main players (groups) in the Oz market, normally you would expect one or two % points between the players and 10% in favour/against would be ringing alarm bells...political polls are a case in point..
The split is quite large indeed.
The numbers of course are changing as more people submit and they are changing figures slightly based on factors such as the FAAA putting out an email. Obviously more Flight attendants answered on that day.
I'm not sure the "disgruntled employee" factor makes much difference. If it is assumed that "disgruntled employees" are more likely to be in a union or read Pprune or respond to a survey, would that not then be the case for all employers involved? By asking the same questions of all sample groups I think we at least will come up with very good cross comparison data.
The graphs I used were just the ones in Survey Monkey and as stated on other thread, they could be better. I may try upgrading the package or yes, someone smarter than I could do some better analysis and produce better graphs.