Marshall, was there any preference among those who were able to distinguish between the samples?
One thing I've taken away from the ExBeeriments is that often there may be a difference between the two beers but which is deemed better is not clear--some like the one, some like the other.
Of the 21 tasters, 12 could distinguish between them, but 9 could not. Of those 12, was there a preference?
***********************
As I've noted before, I love the willingness of people like yourself to experiment and then share the results with others.
I've actually used exbeeriments in my class (statistics) to show the difference between statistically-significant and substantively important. As a suggestion (and again, not being critical, I love that you do this), you might consider adding a preference column or data point or some such that shows whether there is a preference, and whether that preference is statistically-significant. Hard to show with very small samples, but I hearken back to some of what you've done in exbeeriments, where you said you could distinguish between the two beers, but you had no preference finding each delicious.
I'm thinking out loud here a little bit, but it seems common to find a process difference people can detect--BUT, it's not something that makes the beer terrible--just different. And some like the one but not the other.
[I can imagine an exbeeriment where 16 out of 20 could distinguish between the two beers--but 8 prefer the one while 8 prefer the other. What would one take away from that? Me: the process produces a difference but the results depend on what you like. Ergo, no difference.]
When I look at the exbeeriments, I like to examine whether there was a difference that could be detected--AND whether that difference was related to preference. For example, this exbeeriment shows people can distinguish between the beers, but as far as preference....
http://brulosophy.com/2015/09/21/dry-hop-quantity-a-little-vs-a-lot-exbeeriment-results/
Anyway, my sober ramblings for the day.