Capability Analysis (Poisson Distribution) Results - Quick Tab

Select the Quick tab in the Capability analysis (Poisson distribution) results dialog box to access the options described here.

Observed Capability Summary

Click this button to create a summary graph with the following components displayed:

  • U-chart. This chart is useful to determine that, overall, the process is (or is not) in control; see also the Quality Control Help topics for details.
  • Chart of cumulative DPU (defects per unit) across samples. This chart is useful to determine if a sufficient number of samples were taken to estimate a stable value of the percent defective, or the defect rate.
  • A scatterplot of defect rates or percentages vs. sample sizes, if unequal sample sizes were encountered. This chart is useful to determine if the defect rate is independent of the sample sizes.
  • Scatterplot of observed defects vs. expected defects, if fixed sample sizes were specified. This plot is useful to determine if the chosen distribution (binomial or Poisson) is appropriate for the respective data;
  • A distribution histogram of the DPU (defects per unit). This chart will also indicate the target value, if one was specified. This plot is useful to determine if the chosen distribution is appropriate for the respective data, and also allows you to determine the quality of the process relative to the specified upper specification limit.

Summary Statistics

Click the Summary Statistics button to create a spreadsheet containing the following summary descriptive statistics:

  • The Mean DPU (Defects per Unit) value
  • Lower and upper (95%) confidence interval
  • The minimum and maximum DPU values
  •  The target value

For details regarding the computations of these statistics, see the Computational Details.

Runs Tests Summary

Click the Runs Tests Summary button to create a spreadsheet containing the summary results for the runs tests specified on the Capability analysis (Poisson) dialog box - Runs Tests tab.

For details regarding the interpretation of runs tests, see also the Runs Tests topic in the Glossary.

Outliers

Click the Outliers button to create a spreadsheet summarizing the number of samples falling outside the upper and lower control limits of the respective u chart.

Poisson Capability Summary

Click the Poisson Capability Summary button to produce a histogram of the respective frequency distribution of defects (Poisson frequencies).

This graph also shows the fitted Poisson distribution (using the average sample or unit sizes, if unequal sizes were encountered) to the observed frequencies. Based on the fitted distribution and the user-specified upper specification limit (USL), Statistica will also compute the observed and expected PPM (parts per million) defective or non-compliant, and estimate the process capability using the percentage method, consistent with ISO 21747 M1,l=3,d=6).

See Computational Approaches for Process Capability in the Process Capability ISO/DIN (Time dependent distribution model) Overviews.

Note - distribution fitting summary. The fit of the Poisson distributions are evaluated using the Chi-square method (observed vs. expected frequencies); the same methods are used in the Distribution Fitting module (see Discrete Distributions in the Distribution Fitting Help topics).

See also Example 10: Capability Analysis (Binomial), Process Analysis - Capability Analysis Binomial and Poisson, and Capability Analysis - Binomial and Poisson - Computational Details.