Signal and noise
Signal and noise is a theme within reasoning about data. Links to relevant activities and resources are on the right hand side of this page.
It has been proposed that the key organising idea in the study of statistics as an exercise in Exploratory Data Analysis should be the study of noisy processes within which can be detected a signal given sufficient output.
Curriculum materials have been criticised as portraying averages as mere summaries of groups of values, in contrast to the statistician’s perspective on averages as central tendencies or data as ‘a combination of signal and noise’.
A focus on signal and noise commits one to think about the distribution of data around a signal, so that signal plays a key role, in contrast to a view in which the mean (or other measures of average such as mode and median) is simply one of many parameters of the distribution.