Notice
You are now leaving support.minitab.com.
Click Continue to proceed to:
Overall standard deviation is the standard deviation of all the measurements and is an estimate of the overall variation of the process. If your data are collected properly, the overall standard deviation captures all sources of systemic variation. In that case, it represents the actual variation of the process that the customer experiences over time.
If you use a data transformation when you perform the capability analysis, Minitab also calculates StDev(Overall)*, the overall standard deviation of the transformed data.
Compare the within-subgroup standard deviation with the overall standard deviation. A substantial difference between the within-subgroup standard deviation and the overall standard deviation may indicate that the process is not stable, or that your process has other sources of variation in addition to the variation within subgroups. Use a control chart to verify that your process is stable before you perform a capability analysis.
Minitab uses the overall standard deviation to calculate Pp, Ppk, and other measures of the overall capability of your process.
Between standard deviation is an estimate of the variation between the subgroups.
If you use a data transformation when you perform the capability analysis, Minitab also calculates StDev(Between)*, the between-subgroup standard deviation of the transformed data.
Larger values of between-subgroup standard deviation indicate greater variation between subgroups. For example, if each subgroup is collected from a different batch of items, large between-subgroup standard deviation indicates a large amount of variability between the items in different batches.
Minitab uses the between-subgroup variation to calculate the between/within subgroup variation.
Within standard deviation is an estimate of the variation within the subgroups. If your data are collected properly, the within-subgroup variation should not be influenced by changes to process inputs, such as tool wear or different lots of material. In that case, the within standard deviation represents the natural and inherent variation of the process over a short period of time. It indicates the potential variation of the process if shifts and drifts between subgroups were eliminated.
If you use a data transformation when you perform the capability analysis, Minitab also calculates StDev(Within)*, the within subgroup standard deviation of the transformed data.
Larger values of within-subgroup standard deviation indicate greater variation with the subgroups. For example, if each subgroup is collected from a different batch of items, large within-subgroup standard deviation indicates a large amount of variability among the items within each batch.
Minitab uses the within-subgroup standard deviation to calculate the between/within standard deviation.
Between/within standard deviation is a single value that includes both the variation between subgroups and the variation within subgroups. Between/within standard deviation is the square root of the sum of the between-subgroup variance and the within-subgroup variance.
If you use a data transformation when you perform the capability analysis, Minitab also calculates StDev(B/W)*, the between/within subgroup standard deviation of the transformed data.
The between/within standard deviation reflects the total effect of common-cause variation in your process.
Compare the between/within standard deviation with the overall standard deviation. A substantial difference between the between/within standard deviation and the overall standard deviation may indicate that the process is not stable, or that your process has other sources of variation in addition to the variation between/within subgroups. Use a control chart to verify that your process is stable before you perform a capability analysis.
Minitab uses the between/within standard deviation to calculate Cp, Cpk, and other measures of the between/within capability of your process.
You are now leaving support.minitab.com.
Click Continue to proceed to: