An error which varies in an unpredictable manner, in magnitude and sign, when a large number of measurements of the same quantity are made under effectively identical conditions.
Random errors create a characteristic spread of results for any test method and cannot be accounted for by applying corrections.
Random errors are difficult to eliminate but repetition reduces the influences of random errors.
Examples of random errors include errors in pipetting and changes in incubation period. Random errors can be minimized by training, supervision and adherence to standard operating procedures.