Since I have no other information, I'm going to assume you're meaning that H0: π = 0.5. If this isn't true, let me know so I can edit this response.
Going on that, your standard error for the estimate will be √(0.5•0.5/200) ≈ 0.03536. You're using 0.5 and not 0.65 because we assume the null hypothesis is true. Then you divide the difference between 0.65 and the null by this standard error to find your test statistic (t = 4.243). The p-value would be with 199 degrees of freedom, which should be pretty close to normal if you prefer to use that. Either case, the p-value is less than 0.0001, so you're obviously rejecting the null hypothesis.
