If you're serious about grad school, take as many of the basic math courses as you can stand and have time for. Calculus, multivariate calc, linear algebra, probability theory. These will show your analytical capabilities more clearly than most social science stats courses, including intro poli sci methods for PhDs. However, taking a stats class might be useful if you can do it in time to be able to use some stats techniques in your senior thesis. But a solid qualitative thesis plus decent grades in multivariate calc will be enough to put you ahead of that pack in applications. (plus GREs, overall GPA, rec letters, etc., of course)