When we started to build QualStat, we had the goal of building a user friendly statistical software package. Having a multitude of statistical tests that were inaccessible by most users wasn’t acceptable. We focused on usability while providing a powerful list of statistics with an emphasis on multivariate statistical process control. Try out QualStat to see just how quickly you can be productive.
Usability Features
- Data Handling
- Imports data from the clipboard as well as Excel, ODBC, historians and text files
- Exporting of data, reports and charts/graphs
- Calculated fields can be quickly created within the software
- Hiding of individual observations and variables without permanent deletion
- Filtering of observations based on logical expressions
- Grouping of variables with user supplied tags
- Watched Variable list that provides variable values through software
- Graphs and Charts
- Point selection displays information window offering additional data
- Copy and paste the selection of points (i.e. points selected on one chart can be automatically selected on another chart)
- Modify the chart display defaults for the application
- Reporting
- Charts and Analysis reports can be added to a “Work log” that is saved with the data file
- HTML based reporting across the program can be printed and well as exported to other software (ex. MS Word)
- Default report styles can be modified by the user
Statistical Tests and Methods
- General Calculations
- counts, means, minimums, maximum, skewness, kurtosis, standard deviation, variance, covariance, correlations
- calculations can be performed for each value of a grouping variable
- Trend test
- Normality test
- t-Test
- Paired t-Test
- One Way ANOVA test
- Non-Parametric tests
- Wilcoxon Signed Ranks test
- Kolmogorov-Smirnov (KS) test
- Kruskal-Wallis One Way test
- Graph Types
- Histograms
- By Observation trend charts
- XY Charts
- Box and Whisker plots
- 3D Plots
- Chart Matrices (trend charts and xy charts)
- 3D Surface Plots (wireframe, solid and multi-layer)
- Correlograms
- Univariate Statistics Process Control
- Shewhart chart (X bar, R and S charts)
- sub-group based on specific sub-group size or based on variable
- overlay histogram and runs rules
- special printed reports
- Individuals chart
- Moving Average, Moving Range chart
- Exponentially Weight Moving Average chart
- CUSUM chart
- ATT Runs Rules
- Capability Analysis charts
- Shewhart chart (X bar, R and S charts)
- Multivariate Statistics
- Multiple Regression
- Search procedures: stepwise, forward and backward
- Consideration of square, logarithm and product terms without needing to create new variables
- Optional creation of additional data columns for:
- predicted value
- residual value
- residual standard deviation
- standardized residual
- studentized residual
- leverage value
- Cook’s distance
- Ordered residual plot
- Residual by Dependent variable plot
- Combined R square and Standard error plot for Model Search results
- Surface plots for computed models allowing conditional optimization
- XML exporting of models
- Principal Component Analysis (PCA)
- Compute eigenvalues and eigenvectors
- Analysis report for PCAs
- Graph multiple PCA on same chart
- 3D plotting of PCAs
- Use of MVSPC Phase 1 baseline for evaluation
- Contribution graphs
- Hotelling’s T2
- Phase 1 Baseline creation
- Q-Q Plot
- MYT signal decomposition in Phase 1
- Analysis reports in Phase 1
- Phase 2 Evaluation from previous baseline
- MYT signal decomposition support in Phase 2
- Conditional and Unconditional component plots
- Un-squared conditional component plots
- Pareto charts for decomposition
- Analysis reports for Phase 2 evaluation
- Remove mean effect
- Create sub-models
- Support for batch processes
- Batch mean test
- Batch variance test
- Multiple Regression