The issue with (A) is that the argument is not saying that the inference/conclusion is flawed because the survey is conducted by a biased party, i.e. radio station, rather the sample is drawn from a biased population as you have correctly pointed out - radio station listeners are more likely to be pleased with its format, otherwise they would not listen to it in the first place. The radio could still conduct an unbiased survey if it randomly surveyed people from a general population regardless of whether they listen to the radio station or not, so the fact that the radio station is a biased party is not at issue here.
Let me know if this helps and if you have any other questions.