Using the same data - im trying to construct a null

Question

From a sample of 35 presidential elections - 19 of those have been Republican (R) victories - and 16 have been Democrat (D) victories.

1. I need to know how to construction a 95% confidence interval about the proportion of R victories.
2. Using the same data - I'm trying to construct a null hypothesis to show "rejection of the null hypothesis that the true proportion of R victories is greater than what would be produced by flipping a fair coin. So what is the best option for the null hypothesis?
3. What is the rejection region?
4. What is the test statistic?
5. Do we reject or fail to reject the null hypothesis?

All of the above is at 95% confidence level.