The following information may be specified for Variables data when defining Characteristics.
Subgroup Size: The number of data values that will be collected within each rational subgroup. The subgroup size (ss) is inherently linked with the control chart type (X-Bar / Range (1<ss<11), X-Bar / Sigma (1<ss), Individual-X / Moving Range (ss=1), Run Chart (ss=1)), although the software allows selected charts to be used regardless of how the data was collected. The default subgroup size is used for data entry.
High Spec: The Upper Specification Limit (USL), aka Upper Tolerance Limit, usually defined by the internal or external customer or by management. Enter a numeric value, which must be a larger number than the Low Spec (below).
Low Spec: The Lower Specification Limit (LSL), aka Lower Tolerance Limit, usually defined by the internal or external customer or by management. Enter a numeric value, which must be a smaller number than the High Spec (above).
A Note on Specification Limits: Both the Upper and Lower Specification must be provided to calculate Cp and Pp; one or the other or both must be provided to calculate Cpk and Ppk.
Fixed UCL: Enter a numeric value for the user-defined Upper Control Limit, which must be a larger number than the Lower Control Limit (below).
Fixed LCL: Enter a numeric value for the user-defined Lower Control Limit, which must be a smaller number than the Upper Control Limit (above).
A Note on Control Limits: Control Limits must not be confused with Specification Limits, which are defined by customers. The text boxes for Fixed control limits (above) are usually left blank so that the software will calculate the statistical Control Limits based upon the data being analyzed. A typical use of the Fixed Limits option is for entering a ‘historical control limit,’ which is previously calculated based on prior data.
Curve Fit: Select either Johnson or Normal distributions. When Johnson is selected, you may also input the Lower and/or Upper Bound. See also Curve Fitting
Autodrop: Select this option to automatically remove the effect of subgroups violating control limit or run test rules. These subgroups are still shown on the chart, but their statistical bias is removed from the calculation of the control limits and process sigma.