Adding onto to the previous, great answer:

Another reason B is wrong is because the argument that we're trying to weaken is that:

The audience members surveyed, who tended to think Lopez had the better argument, were inherently biased to favor Lopez to begin with.

Even if you skipped over the sample population issue of live audience members surveyed vs all debate viewers surveyed, answer choice B tends to conflict with the premise, if you can equate "better arguments" to equal "more persuasive". It definitely does not weaken the argument and I could see how one could argue that it strengthens the argument, if Answer Choice B is saying that a smaller subset of live audience members could see Tanner as more persuasive. Thus, the non-live audience members could be inherently biased for Lopez and the argument still stands.

D is the correct answer because it directly addresses the core assumption in the argument that the audience members were inherently biased in favor of Lopez before the debate began. D flips this on its head, saying that they were instead inherently biased of Tanner, which would absolutely weaken the argument.