My exciting committee meeting isn't until tomorrow (Tuesday), but I spent a lot of time in the lab this evening and things are suddenly looking a lot better.
I've t-tested and ANOVAed and Levene'd and Welched it; I've standard deviationed it and meaned it; and suddenly my results are looking a lot more promising. The key was to remove the two outliers more than two standard deviations from the mean.
Oh, the results aren't clean at all. No, there's one experiment that's just come out wacky and I still have no idea why, but everything else is tending towards my predictions. Tending means usually there's a 85-90% chance I'm right. The random threshold in the social sciences is that there has to be a 95% chance you're right, and it's pretty categorical. 95% chance = publish and get a job! 94% chance = education is its own reward!
Anyway, I've now gone from having almost no idea what I wanted to say to my committee other than "help!" to having a pretty clear idea what the three next best steps would be. So I'm happier now.