Originally Posted by
entropy
you really think 5-6 reports is anything close to ... statistically significant?
there's always reporting bias in both IBB's and self-reported 'stats'
its just about as good as anecdotal evidence.
In any study, small subgroups will be... small. Doesn't mean you can't say useful things about them. I used to run into this issue when working on neuroimaging studies, where subgroup sizes are sometimes that small. Some people just won't believe you even if you have the analysis that shows you have the necessary statistical power.
In any case, the overall data is supported by more than 600 reports, which is a lot better than scattered posts on FT, if you ask me.
Originally Posted by
mitchmu
I suggest it has gotten insanely worse for some people and if you slice by airport combinations the data will show it.
[Edited to remove my confusion about this quote.]
In order for the averages to come out the way they have, it would have had to get less worse for some other group of people.
Incidentally, silver and gold have hovered around 30% and plat and 1K have hovered around 60% since the beginning of the project.