Bias, values and scientific knowledge: Towards a functionalist interpretation

The purpose of scientific study has always been to produce reliable knowledge about the phenomena that make up reality. To this end, the process of scientific inquiry developed throughout the history of the scientific method with the singular goal of objectivity and strict adherence to information inferred from real data. The objectivity of modern science makes it the most reliable source of knowledge regarding a wide range of aspects of society and of each of our private lives. In spite of this need for objectivity, the institutions that build the scientific community and the people that produce scientific knowledge are still human and thus vulnerable to human errors, both isolated and systemic ones. The purpose of this dissertation is to describe and analyze some of the existing conceptualizations of bias, and to discuss the influence that bias still has over the production of scientific knowledge and what information on the nature of bias can be inferred from these influences. the discussion of this concept will begin with the analysis of some models of bias, and then move onto a discussion of the influence that biases have on the methodology and on the practice of the scientific community, with a particular focus on the issues of replicability and publication biases, and on the ways these biases emerge from compounding social incentives at various levels of the institutions that make up the scientific community. In the second half of the dissertation, the focus will broaden slightly to discuss the differences between biases and other influences produced by non-epistemic values that can impact the production of scientific knowledge. Finally, as a concluding note , we will give three practical examples of the influences of bias on various fields of scientific inquiry, providing good examples of the complexity of the issues created by the presence of bias in the scientific field and of how these issues require a multifaceted and nuanced approach in order to be solved.