Statistical Errors

Abstract

Statistical errors, also called uncertainties, can be divided into four types.

• 1. Instrumental errors. These are caused by lamp instability, noise in the photosensor, and resistance fluctuation in electric circuits such as the power-supply units and photoelectric assembly.

• 2. Reading errors. These occur because the operator is not perfect in reading a fluctuating needle or digital display. In the case of on-line processing of measuring values with the help of a computer, or of print-out of measuring values, this type of error does not occur.

• 3. Stage-object errors. These arise because of natural variations in the stage object or else of artificial variations introduced by the preparation process. The first may in some cases be utilised in the characterisation of the stage object; the second must just be minimised by careful preparation.

• 4. Errors in repetition measurements. These arise whenever a module has to be re-set between measurements; they have to be minimised by reducing, as far as possible, the operations of re-setting and by using fiduciary marks or stops as aids in re-setting.