The evaluation of the uncertainty in knowing a directly measured quantity

This paper addresses the problem of evaluating the uncertainty in knowing a quantity measured directly with a measuring instrument. It is shown that, under some conditions, the uncertainty associated with the uncorrected result can be obtained from that associated with the correction. For this, the number of calibration measurements and the uncertainty in knowing the value of the reference standard must be known by the user of the instrument. The paper also discusses the evaluation of the uncertainty in case of indications corrected by an average of correction values over a given measurement range, and in case of a correction of null value.