Knowledge and Truth in Data Science: Theory Without Theory?

Abstract

Essay contemplation on whether Data Science with its correlation without explanation, in particular without need of cause and effect explanation, normatively ought to be the new scientific method. This essay argues it should. Not only is statistical correlation the next step in scientific progress for its epistemology necessary to deal with the huge amount of information and data that is now available and that will increase as Technological Society progresses, but it also provides a predictive and pragmatic methodology that avoids the illogic and paradoxes of cause and effect and induction. Furthermore, statistical correlation has a potential for allowing morality and ethics to logically and structurally avoid Hume's Guillotine.

Keywords: data science, correlation, scientific methodology, statistics, normative, induction, cause and effect

JEL Classification: Y1, C40