Since analytical chemistry is the science of making quantitative measurements, it is important that raw data is manipulated and reported correctly to give a realistic estimate of the uncertainty in a result.
Simple data manipulations may only require keeping track of significant figures . More complicated calculations require propagation-of-error methods.
The uncertainity in a result can be categorized into random error and systematic error.
See the statistical formula document for more quantitative descriptions of describing and testing data sets.
