Measurement and Uncertainty

Abstract

Performing analytics requires more than a superficial understanding of terminology. This chapter provides a conceptual overview of stochastic variability of data, distributions of data, the progression of data into information, then knowledge, then wisdom, and statistical error. Data are derived from existing databases and extracted into spreadsheets which contain datasets. Most statistical software is designed to manipulate numerical values; “R” has a structure called “dataframe” which can handle both numerical and nominal (text) values. Converting data into useful models requires an understanding of the concepts of probability and odds, as these are the underpinnings of mathematical modeling. Key concepts of working with numbers in databases are summarized, including parameters, variability, and the difference between observed and expected (the residual).