Should the repeatability of the instrument under test be included in test uncertainty?

Published

August 18, 2017

Author(s)

Craig M. Shakarji, Steven D. Phillips

Abstract

Should the repeatability of the instrument under test be included in test uncertainty? The answer— which surprises many—is no, at least for the range of cases documented in this paper. This has been an area of confusion even among experts around the world. Some ramifications of this confusion have been known cases where companies have had to greatly and artificially inflate their accuracy specifications for their measuring instruments in order to meet decision rule requirements for acceptance testing. The purposes of this paper are to (1) Clearly identify a large set of common instrument testing/calibration situations where instruments are verified to their accuracy specifications where the repeatability of the instrument under test should not be included, (2) Take the reader through a detailed tutorial path to gain understanding as to why it must be that such a component should not be included in those cases, and (3) Identify four common mistakes that lead even experts to incorrectly include the repeatability component: These being (i) misunderstanding that there are two measurands involved in testing, (ii) misunderstanding the difference between the test value uncertainty and uncertainty in the overall test, (iii) misunderstanding the role of test value uncertainty by envisioning its use in subsequent measurements made by the tested instrument, and (iv) misapplying the inclusion of the repeatability of the calibrating system performing the test to include the instrument under test.

Proceedings Title

NCSLi 2017 Workshop & Symposium

Conference Dates

August 15-17, 2017

Conference Location

National Harbor, MD

Pub Type

Conferences

Download Paper

Local Download

Keywords

calibration, repeatability, test uncertainty, uncertainty

Citation

Shakarji, C. and Phillips, S. (2017), Should the repeatability of the instrument under test be included in test uncertainty?, NCSLi 2017 Workshop & Symposium, National Harbor, MD, [online], https://tsapps.nist.gov/publication/get_pdf.cfm?pub_id=924082 (Accessed April 19, 2024)