Uncertainty is not measured but is estimated. Uncertainty estimations are developed by very detailed mathematical analysis and by observations. Uncertainty and its determination is a science all of its own and is often very difficult, complicated and time consuming. Many calibration labs employ full-time people just to determine uncertainty of measurement results obtained within their facility.
NIST recognizes two main methods of obtaining uncertainty:
Method A - This method is determined by analysis, measurements and observations of results. All the contributing elements that give rise to any error in measurement are measured and the standard deviation obtained. The measurement results of all the contributing components are then added together using the square root of the sum of the squares and an overall standard deviation determined. The overall uncertainty is then stated as either two times the standard deviation or three times the standard deviation.
Method B - This method is determined of the uncertainty by a theoretical and mathematical analysis, rather than by observation.
Honeywell carries out all uncertainty measurements using Method A.