This is background context describing what political candidates normally do. It's not something the author is trying to prove. Nothing in the argument supports the claim that candidates normally send out material to influence opinion. The author simply states it as a known fact to set up the contrast with what Ebsen's campaign did.

This is a premise, not the conclusion. The author treats this as an established fact and uses it to support her inference about the purpose of the ads. If you think (B) is the conclusion, ask yourself: what other line in the argument is offered to prove that the ads were sent to too few households? None. The author just asserts it. That's a sign it's not a conclusion. The author never tries to convince us that the ads went to too few people. She takes that for granted and asks: given that, what were they for?

This is the conclusion. The word "evidently" signals that the author is drawing an inference. She doesn't know for certain that the ads were a test. She's piecing together the evidence (too few households, wide variety of topics, heavy spending on follow-up surveys) and concluding that the most plausible explanation is that the campaign was testing the ads' effectiveness.

This is a premise that supports the conclusion. The wide variety of topics is consistent with a test, because you'd want to try different messages to see which ones resonate. But nothing in the argument is offered to prove that the ads covered a wide variety of topics. The author just states it as a fact.

This is a premise that supports the conclusion. Spending heavily on follow-up surveys to measure the ads' effect is exactly the kind of behavior you'd expect from a campaign running a test. But, like the other premises, nothing in the argument is offered to prove that the campaign has been spending heavily on follow-up. The author states it as a fact and uses it to support her inference.