>I know almost nothing about statistics. Is this the right way to do a poll?
Depends on whether you want to do statistical analysis on your survey results, or just post percentages on your social media outlet to btfo the libs.
From what you've posted it looks like the latter, in which case the only statistical advice I can offer is that you'll probably want to stick to yes-no questions as you're currently doing. (Except that I'd suggest changing Q4 to "Are you planning to vote for President Trump in 2020? (Y/N)", since the respondent may be independent, abstaining from voting, etc.)
Also, while I'm not going to comment on your politics (not American), you should at least get a neutral third party to look over the phrasing of your questions (I seriously hope you're not expecting to get an unbiased response for Q6). For example, do you really need the adjectives "thriving" and "radical" for Q2?
If you actually want statisticians to take your poll seriously, you'll have to at least fix the above. After that, I'd consider replacing the binary questions with a range of values where possible (look up "Likert scale" on Wikipedia to get an idea of what I'm talking about). E.g. for Q1, even asking for their opinion of socialism on a 5-point scale (say) would allow you to differentiate the moderates from the extremists (on either side).
Even better, you could split up some of the questions to allow for more nuance (e.g. for Q5 the respondent could feel that the govt should be less involved in terms of providing coverage, but more involved in regulating costs). Or, if the survey format permits, add a "text box" for open-ended responses so that the respondent can clarify their choices (this will help in outlier detection).
I'm sorry, but as it stands this the kind of pool data analysts refer to when they say "garbage in, garbage out". If you're serious about polling "the right way", I'd strongly suggest getting professional involvement rather than winging it on your own.
