- P-ISSN 1225-0163
- E-ISSN 2288-8985
New approach is developed employing the overall uncertainty obtained from multiple measurements to evaluate the statistical significance for the difference between a given reference value and its measured value determined in a lab. The overall uncertainty is determined by separate combinations of the uncertainties arising from systematic and random effects. It is shown that the uncertainty term in regular t-test can be underestimated by n measurements.