Quality
Quality engineering alongside the rest of your statistics
Monitor a process, assess its measurement system and capability, and support acceptance and reliability decisions, in the same project as your data preparation, models and experiments.
Measurement standards you control
Measurement-system studies follow the AIAG MSA 4th editionstandard, including its spread multipliers, total-variation basis, ndc and guidance bands. Your company can also define its own named profile with different settings.
Results that keep their settings
Every saved study records the standard, its version and the effective settings. Reopen it on another machine and it uses the recorded configuration, labelled as such, instead of silently switching to whatever default is installed.
Statistical process control
Examine variation over time or run order, then turn a chart into a documented decision.
- Individuals and moving-range charts
- Subgroup charts for means and dispersion
- p, np, c and u charts with binomial and Poisson limits, including unequal sample sizes
- Laney p′ and u′ charts for overdispersion
- Rare-event charts based on opportunities and elapsed time
- CUSUM and EWMA charts for small sustained shifts
- Finite-window moving-average charts
- Short-run standardized Z–MR charts across mixed products
- Multivariate subgroup monitoring with Hotelling’s T² and joint-spread charts
- Multivariate EWMA
- Nelson sequence rules 1–8, each switched on or off, with Rules 1–4 on by default
- Approved reference limits, process stages and recorded responses to signals
Measurement system analysis
Understand how much variation comes from the measurement process before you judge the process itself.
- Randomized collection worksheets for planned studies
- Crossed Gage R&R by ANOVA, average and range, and range approximation
- Studies with an additional site or fixture factor
- Nested and destructive studies
- Linearity and bias, Type 1 (single standard) and stability studies
- Preliminary repeatability screening
- Attribute agreement with exact agreement and kappa, including ordinal studies
- Binary gage response curves from counted trials
- Method comparison by Deming regression
Specifications and capability
Relate a stable process to its specifications, with the uncertainty of each estimate kept visible.
- Within-process Cp and Cpk and overall Pp and Ppk, with confidence intervals
- Non-normal capability using percentile indices, including Johnson transformations
- Defective-unit (binomial) and defect (Poisson) capability
- Within- and between-batch capability with a variance-components model
Acceptance, tolerance and reliability
Support inspection, sampling and life-data decisions with explicit assumptions.
- Attribute acceptance plans from defective-unit counts
- Acceptance sampling by variables with its operating characteristic
- Tolerance intervals covering a stated population fraction
- Right-censored life distributions and mixed exact and interval-censored life data
- Accelerated life models at declared stress conditions, including Arrhenius
- Reliability demonstration and precision test plans
Guided studies
Step-by-step workflows that keep the reasoning with the result.
- A guided measurement-system study, where the physical design determines the analysis
- From measurement to stability and capability: choose the chart, establish the Phase I population and approve limits
- Calculated capability kept separate from a qualification decision
Checked against published references
Selected control-chart, capability, measurement and acceptance-sampling calculations are part of the reference assessment.