What an interesting distribution... From the table it looks like there is a correlation between method and results. Are people less willing to put their trust i

What an interesting distribution... From the table it looks like there is a correlation between method and results. Are people less willing to put their trust in Trump over the phone than in anonymity, or are the people who are making calls biased, or...? :)