Uncertainty Calculator (rev. 3.2 released March 2002)

A FREEWARE application produced by Chris Grachanen of Hewlett-Packard (formerly Compaq) for Windows® 2000 (inc. Win98, NT, etc.). The original program which has a more modest, but still very useful, specification is also available for Windows® 3.1x.

Uncertainty Calculator has been developed to address the need to document analyses of measurement accuracy (uncertainty) in a simple, straightforward manner and is consistent with the ISO Guide to Uncertainty in Measurement ('GUM'). It includes examples from authoritative documents such as those published by NIST, NCSL, UKAS, SINGLAS and EA.

Key features (not in the original program) include:

• Effective Degrees of Freedom – The Welch-Satterthwaite equation for computing effective degrees of freedom has been incorporated with the ability to define group and individual Type ‘B’ relative uncertainties.
• Numerical Methods for Computing Sensitivity Coefficients – Several Type ‘A’ and Type ‘B’ algorithms have been incorporated for computing sensitivity coefficients including Type ‘B’ uncertainties derived experimentally. Common partial differential equation computations.
• Correlated Input Quantities – Methodology for computing correlated input quantities.
• System File – Provides the means for defining application paths and for assignment of a default data file directory. The System file is also used to assign a report application (Microsoft Write® , MS Word® or Excel® ) and for entry of default report header information (e.g. company name, department, etc.).
• Equation Writer – Provides a handy means for creating and documenting measurement equations.

Features and enhancements included in UnCal3 are:

· Fields for entering individual and expanded uncertainty units
· Fields for entering correlation coefficient approximations for each uncertainty value
· Tab menu for accessing all major functions / screens
· Student t distribution divisors expanded beyond 'GUM' table
· Normal distribution divisor selections have been added
· A worksheet for preprocessing uncertainty components comprised of formulae
· A worksheet for deriving sample statistics from measurement data
· Computation tool for determining an uncertainty value from a given Standard Uncertainty
· Bayesian method for computing Effective Degrees of Freedom
· Type B individual Effective Degrees of Freedom computations
· Quick Print command for printing data without need for Windows application
· Excel tools for importing spreadsheet contents
· Individual Type A and Type B uncertainty components increased from 10 to 20 each
· Expanded on-line dB / Percentage Calculator with memory function
· Options screen for changing default setups
· Optimized for 800 x 600 display with auto resizing for higher resolutions
· Quick print command for displaying contents of main screen
· Global column change(s) tool
· On-line tutorial
· Hot Key assignments for commonly used functions
· Manual or Auto-sequence data entry
· Remarks expanded from 20 to 40 lines
· Downward compatible with previous revisions

What Changed since rev.3.1?
Bug fixed with computing negative correlations coefficients (negative correlations sign was not carried through the calculations such that negative correlations were calculated as positive correlations). A new option to normalize Type A and B Standard Uncertainty & Variance RMS results to 1 sigma (does not affect Type A and B Combined RMS results.). New Sample Statistic functions that computes Standard Error of the Means and Coefficient of Variation.

®Microsoft Corporation

Opinion

Great stuff !  The only negative thing I can think of is that, sadly, you do have to know the source and value of your potential measurement errors.  If only Chris could just figure a program to solve that little problem....!

But seriously, users will find UnCal3 to be a "professional" grade application that is not bettered by those offered commercially . Chris's decision to continue its "freeware" status with this major revision is an outstanding contribution to the global measurement community and richly deserves grateful acknowledgment.

Disclaimer

Agilent Technologies is not responsible for the content or support of any material provided by a third-party.

Contact the author:

• Christopher Grachanen