In the 1936 presidential election Franklin D. Roosevelt defeated Alfred E. Landon in a landslide vote.  A Landon victory had been predicted by the Literary Digest, a magazine that ran the oldest, largest, and most widely publicized of the polls at the time.  The Literary Digest's final prediction was based on 10 million sample ballots mailed to prospective voters, and 2.3 million were returned.  The sample of voters was drawn from lists of automobile and telephone owners.  Despite the massive size of this sample, it failed to predict a Roosevelt victory, being off the mark by 19%.  The Literary Digest was wrong because:

a) the sample size, athough large, was not large enough

b) the right research questions were not asked

c) respondents intentionally lied about their preffered candidate

d) the sample used was not representative of the actual population at the time