Statistics is a browser-based collection of tools for data analysis, quality control and process improvement.
The Statistics tool is continuously developed and expanded with new analyses and features. We welcome user feedback — comments, problem reports and suggestions for improvement help us determine the direction of its further development.
It can perform statistical tests, create charts, evaluate process stability and capability, analyse measurement systems and support the design of experiments. No additional software needs to be installed.
The user works in a data worksheet, selects an appropriate analysis, assigns the required columns and receives results as calculations, tables and charts.
What is the Statistics tool?
Statistics combines a data worksheet with analyses available from a single menu. Data can be entered manually, pasted from a spreadsheet or loaded from prepared sample datasets.
The available tools help answer questions such as:
- Is the difference between groups statistically significant?
- Are the variables related?
- Is the process stable over time?
- Is the process capable of meeting specification requirements?
- Is the measurement system sufficiently reliable?
- Which factors actually influence the result?
- How can the distribution and variation of data be presented graphically?
How do you start an analysis?
After opening the tool, the user enters data into the worksheet. Columns can contain numerical values, categories, group identifiers, process stages or dates.
The next step is to select an analysis and assign the relevant columns to the required fields. Depending on the tool, the user can also:
- group the data,
- apply filters and select the analysed range,
- enter lower and upper specification limits,
- add reference lines to a chart.
Prepared sample datasets are available for many analyses. They make it possible to explore the application without first preparing your own data.
Stochastic data
Data can also be created with the random data generator. The generator allows the user to specify the number of observations and distribution parameters, and then write the generated values directly to the worksheet.
The available distributions are:
- normal,
- uniform,
- log-normal,
- exponential,
- Weibull,
- binomial,
- Poisson.
The stochastic data generator can support learning, preparing exercises, testing analyses and simulating different types of process behaviour.
Statistical tests
The module contains tests and models for describing data, comparing groups and examining relationships between variables.
The available methods include:
- descriptive statistics and graphical summary,
- normality test,
- two-sample t-Test,
- paired t-Test for before-and-after measurements,
- analysis of variance — ANOVA and residual analysis,
- equal variance test and Chi-Square test,
- two proportions test and two-sample Poisson rate test,
- correlation and linear, quadratic, multiple and stepwise regression,
- sample size calculation.
These analyses can compare machines, suppliers, technologies, production shifts and results obtained before and after a process improvement.
Charts and graphical analyses
Graphical analyses reveal the distribution, sequence and variation of data. They often expose patterns that cannot be seen in a results table alone.
The application includes:
- time series plot,
- histogram,
- boxplot,
- interval plot,
- dot plot,
- Pareto chart,
- Multi-Vari chart.
Data can be grouped by machine, product, operator, supplier, shift or another category. Filters can limit the analysis to a selected range or subset of data.
SPC control charts and process capability
SPC control charts help distinguish natural process variation from signals indicating that a special cause may have occurred.
The available tools are:
- I-MR chart,
- I-MR chart with process stages,
- Xbar-R chart,
- Xbar-S chart,
- p chart,
- np chart,
- c chart,
- u chart,
- an interactive guide for selecting the right chart,
- process capability analysis based on Cp and Cpk indices.
SPC supports routine production monitoring, evaluation of process changes and detection of situations in which a process is no longer stable.
Measurement System Analysis — MSA
Process analysis can only be trusted when the measurements used in it are reliable. Statistics includes the following measurement system analysis tools:
- MSA Type 1,
- Gage R&R,
- Attributive MSA.
These tools assess systematic measurement error and instrument repeatability, the influence of operators and agreement between inspectors for categorical results such as “pass” and “fail”.
Design and analysis of experiments — DoE
Design of Experiments tools help plan trials and determine which factors affect a process result. The module can generate a full factorial design and analyse collected results using regression.
DoE can optimise process parameters, reduce the number of costly trials and identify settings that provide the expected result.
Learning statistics with examples
The interactive educational materials explain:
- how to build and interpret a boxplot,
- how the number of bins affects a histogram,
- the relationship between a histogram and an SPC chart,
- how the Central Limit Theorem works.
Prepared datasets allow users to run selected analyses step by step without first creating their own worksheet. Available examples include:
- basic statistics,
- process capability,
- graphical analyses,
- large datasets,
- SPC control charts,
- analysis of variance — ANOVA,
- two-sample t-Test,
- paired t-Test,
- Pareto analysis,
- Gage R&R,
- Attributive MSA,
- MSA Type 1,
- correlation,
- linear regression,
- quadratic regression,
- multiple and stepwise regression,
- equal variance test,
- two proportions test,
- Chi-Square test,
- two-sample Poisson rate test,
- Multi-Vari analysis,
- DoE experiments,
- boxplot.
The datasets include data and guidance for assigning columns to the selected analysis. They can support self-study, training courses and practical workshops.
Who is Statistics intended for?
The tool can be useful for:
- quality and process engineers,
- Lean and Six Sigma specialists,
- manufacturing engineers and production employees,
- people managing process-improvement projects,
- students and training participants,
- users who need to analyse data without extensive statistical software.
The available methods cover basic data analysis and tools used for:
- Six Sigma Green Belt projects,
- Six Sigma Black Belt projects,
- process monitoring with SPC,
- measurement system evaluation with MSA,
- design and analysis of experiments with DoE.