No; the sample is biased.
Work Step by Step
This is poor statistical analysis for a variety of reasons. First, many political radio stations attract an audience favoring a particular candidate, so the population is not truly random. Second, the sample may not be large enough to provide a definitive result (though the first reason is the answer the book is looking for).