have a 24 variation multivariate test running on my squeeze page... traffic is affiliates and paid traffic.
read a post here that had a "hijacked" fb likes code.... where you can iframe any other sites fb like thing... so i decided to add this into my page and test.
one of the things i'm testing is googles fb likes (3.3 mill), askmen(10k) and no likes.
figured thats a good variety.... a ton, a bit and none.
here's the current stats... i mean its still really early but still.

How did you have the like button/box positioned? Was it close to the call of action, above the submit button, in the sidebar, etc?
What were you considering a conversion in this test? a click-through or an actual conversion/lead?
Also, try this fb like test separately rather than in a 24 variation. My experience using a legit fb like button (was for the actual site, not an inflated one) increased conversions enough to keep it (plus got to add fans for free while increasing CR).
it's under the submit button.
a conversion is someone opting in.
why test it separately? i don't get that thinking... its being tested with every other element on the page. thats the only way you're gonna know what works.
how do you read the chart? which is winning?
that just shows the conversions for the FB likes section... right now its basically the same.