Error Source Worksheet Features 
The Error Source Worksheet can be used instead of entering data directly in the Root Variables Data table of the Multivariate Parameter Worksheet.  It is a tool for estimating the measurement uncertainty for a selected variable or error source.  Key features are described below. 

Error Source Description
This box of the Worksheet allows you to enter additional descriptive information about the root variable error.  If no information is entered into this box, then root variable name is displayed by default.
 
Reference Values
This section of the Error Source Worksheet provides a means for establishing the measurement area, nominal units and uncertainty units. If appropriate, a nominal value and other values that characterize the root variable may be entered in this section. If a sample of data has been entered in the Type A Uncertainty section, the mean value of this sample is Included as a reference value. Note that this value is the corrected mean that accounts for adjustments entered in the Mean Correction box.

Type A Uncertainty

Measurement uncertainty for variable or error source can be estimated statistically from Measured Values or Deviations from Nominal that are entered in the Type A Uncertainty data entry table.  Entering values in the table causes various statistics to be computed.
 
The Mean Value is the average or mean of the measured values. The Mean Value may need to be adjusted to account for environment or other factors (e.g., length correction due to thermal expansion). The factor to apply in making this adjustment is entered in the Mean Correction box. If the Error Source Worksheet is called from the User Defined Error Worksheet, the resulting value is the Adjusted Mean, displayed in the Root Variables Data table. 
 
The Mean Deviation is the average or mean of the measured deviations from nominal. If the root variable has a nominal value equal to zero, then the Mean Deviation and the Mean Value are equal. 
 
The Sample Size is the number of values entered in the data table. 
 
The Standard Uncertainty is the computed standard deviation of the data sample.  If you enter a Confidence Level (%) or Coverage Factor an Expanded Uncertainty will be computed.
 
The Use Mean Value Uncertainty option is checked if the standard uncertainty is to be reported for the mean value. If this option is unchecked, then the computed standard uncertainty is used for the Type A uncertainty on the root variable. If this option is checked then the standard uncertainty of the mean is used.

Type B Uncertainty

Measurement uncertainty for a variable or error source may be estimated heuristically from bounding values, or ± Limits, that are expected to contain the error with some specified probability or Confidence Level. If appropriate, a Coverage Factor can be entered instead of a confidence level.  The ± Limits can be specified as a Fixed value, % of Nominal, % of Ref 1, % of Ref 2, or any combination of these quantities. If the Linear Spec Combination is selected, the ± Limits are computed as a linear sum of the displayed values. If the RSS Spec Combination is selected, the ± Limits are computed as a root-sum-square combination of the displayed values.
 
In Type B uncertainty analysis errors can typically be assumed to be normally distributed. However, if a Confidence Level of 100% is entered, then the uniform distribution is applied.
 
Uncertainty Sidekick Pro assumes infinite degrees of freedom for the Type B uncertainty estimate unless otherwise specified.  Since, the degrees of freedom variable quantifies the amount of knowledge available for making the uncertainty estimate, an infinite degrees of freedom signifies complete certainty, i.e., zero uncertainty.  In most cases, it would not be realistic to assume infinite knowledge about the uncertainty.  The Type B Degrees of Freedom Calculator, which is accessed by clicking the Degrees of Freedom button, can be used to provide additional information about the uncertainty in the ± limits and associated confidence level. In essence, it provides a means for incorporating a more realistic assessment of the "fuzziness" or lack of knowledge in an estimate. 

Combined Uncertainty and Degrees of Freedom

The bottom section of the Error Source Worksheet displays the Combined Uncertainty for the error under consideration. The combined uncertainty is obtained by combining the Type A and Type B uncertainty estimates.  The analysis category (Type A, Type B or Type A,B) for the combined uncertainty estimate is also displayed.  The degrees of freedom for the combined uncertainty is computed using the Welch-Satterthwaite formula.