21 Quality in statistics

It seems to be obvious that statistics is a strictly quantitative discipline. However, that is not so, as we shall explain.

Certainly, statistics is a way of arriving at an understanding of the world using techniques for analysing numerical quantities, either measured or counted. ‘Numerical detective work’ is the way the great US statistician John Tukey described statistical analysis in his renowned book Exploratory Data Analysis.

Before the 19th century, statistics was literally ‘state‐istics’, that is, a description of the state (i.e. the nation) – a description which, moreover, focused heavily on qualitative (i.e. non‐numerical) analysis. Questions about a country’s productivity, wealth and well‐being were answered by analyses based on observed characteristics (without necessarily including any measurements), such as its progress in agriculture and industry and its accomplishments in the arts and architecture. An interesting historical essay by de Bruyn (2004) illustrates how this worked in practice.

Do qualitative analyses still have a place in modern statistics? Indeed, they do. Beginners in statistics may form the impression that it concerns itself only with quantitative data and quantitative analyses. However, qualitative data and qualitative analyses are a vital part of statistics, too.

What exactly do the terms ‘quantitative’ and ‘qualitative’ mean in this context? Dictionaries usually define these words by referring back to the terms ...

Get A Panorama of Statistics now with O’Reilly online learning.

O’Reilly members experience live online training, plus books, videos, and digital content from 200+ publishers.