In contemporary epistemology, there are two broad approaches to conceptualizing belief. According to the qualitative conception, belief is all or nothing; an agent either believes some proposition, or she does not. This is often referred to as full belief. Alternatively, according to the quantitative conception, belief comes by degrees; an agent can partly believe a proposition, having a lower or higher degree of confidence in its truth. This is often referred to as partial belief. In my dissertation, I explore the connection between full and partial belief. In chapters 2 and 3, I focus on a paradox that arises when we try to understand full belief as having a sufficient degree of partial belief, namely the Lottery Paradox. Briefly, in a large enough lottery, my degree of belief that my ticket will lose is sufficient for full belief that it will lose. The same can be said for all of the tickets. This puts me in the awkward position of believing of each ticket that it will lose, but also believing that there is a winning ticket. In Chapter 2 I examine a formal system, Ranking Theory, that models both full and partial belief while avoiding the Lottery Paradox. I argue, however, that in doing so it is an implausible model of our beliefs about fair lotteries. In Chapter 3 I show how we can avoid the Lottery Paradox by treating belief as contrastive–what we believe depends on what we are considering as alternatives. Turning from formal to applied epistemology, in chapters 4 and 5 I examine how the relation between full and partial belief has practical consequences in two areas–the use of statistical evidence in legal contexts, and the epistemic evaluation of conspiracy theories. Recently some have argued that the use of statistical evidence in legal contexts shows that we cannot formally reduce full belief to partial belief. Using the contrastive method introduced in Chapter 3, I demonstrate that this is not correct. Continuing with the idea of contrastive belief, in Chapter 5, I examine the epistemology of conspiracy theories through a contrastive Bayesian lens.