The discussion revolves around the concepts of uncertainty in measurements, specifically focusing on systematic and random uncertainties. Participants are exploring the definitions and relationships between absolute uncertainty, systematic error, and random error.
Some participants have provided clarifications regarding the definitions of uncertainties and expressed their views on the terminology. There appears to be an ongoing exploration of the concepts without a clear consensus on the definitions or relationships.
There are indications of varying interpretations of uncertainty concepts, particularly regarding the distinction between systematic uncertainty and systematic error. The discussion also touches on the implications of these uncertainties in the context of repeated versus varied measurements.
BvU said:Any particular reason ?
Thank you for your replies @BvU and @haruspex!haruspex said:As far as I am aware, absolute uncertainty means the absolute amount by which the measured value may differ from the actual value. This is as opposed to fractional uncertainty, which is absolute uncertainty divided by the measured value.
And I find "systematic uncertainty" conceptually awkward. Systematic error is the more usual expression.
So I would say that total error is systematic + random, where each of those may be (consistently) interpreted as fractional or absolute.
That is with regard to repeated measurements which are in principle of the same quantity. If they are for different quantities (because some parameter is being varied) these errors may vary in different ways. E.g. the systematic fractional error my remain constant, while for random error it is the absolute error that is constant.