Abstract

This thesis covers the uncertainty of empirical prediction. As opposed to objectivity, I will discuss the practicality of statistics. Practicality defined as "useful" in an unbiased sense, in relation to something in the external world that we care about. We want our model of prediction to give us unbiased inference whilst also being able to speak about something we care about. For the reasons explained, the inherent uncertainty of statistics undermines the unbiased inference for many methods. Bayesian Statistics, by valuing hypotheses is more plausible but ultimately cannot arrive at an unbiased inference. I posit the value theory of money as a concept that might be able to allow us to derive unbiased inferences from while still being something we care about. However, money is of instrumental value, ultimately being worth less than an object of “transcendental value.” Which I define as something that is worth more than money since money’s purpose is to help us achieve “transcendental value” under the value theory. Ultimately, as long as an individual has faith in a given hypothesis it will be worth more than any hypothesis valued with money. From there we undermine statistic’s practicality as it seems as though without the concept of money we have no manner of valuing hypotheses unbiasedly, and uncertainty undermines the “objective” inferences we might have been able to make.