Publication
Robust statistics has multiple goals, which are not always aligned. Robust thinking grew out of data analysis and the realisation that empirical evidence is at times supported merely by one or a few observations. The paper examines the outgrowth from this criticism of the statistical method over the last few decades.