Measurement
2 tools in the Quality area.
Gauge R&R
Measurement systems analysis after the AIAG MSA Reference Manual (4th edition): a crossed gauge repeatability and reproducibility study by the Average and Range method (equipment, appraiser and part variation from R̿, X̄diff and Rp with the K1, K2, K3 constants of the d2* table, %EV, %AV, %GRR, %PV, %Tolerance and the number of distinct categories, with the X̄ and R charts by appraiser) or by two-way ANOVA with the appraiser × part interaction (kept or pooled at the 0.25 rule, variance components and %Contribution), an attribute agreement analysis (within-appraiser, between-appraiser and versus-standard agreement with binomial intervals, Fleiss' and Cohen's κ, AIAG effectiveness, miss and false-alarm rates), a linearity and bias study (bias at each reference with its t test, the regression of bias on the reference value with confidence bands, the linearity as a share of the process variation, and the independent-sample bias study with the d2* repeatability estimate) and a Type 1 gauge study (Cg, Cgk, %Var). The default is the manual's own three-appraiser, ten-part, three-trial example, whose report (EV 0.2019, AV 0.2296, GRR 0.3058, %GRR 26.7 %, ndc 5) and ANOVA (%Study Variation 27.9 %, ndc 4) the page reproduces; the linearity, bias and attribute boxes carry the manual's linearity and bias examples and Fleiss' worked table.
1 worked example reproduced, listed on the tool page.
Uncertainty Budget
The calibration-laboratory uncertainty budget in the layout of EA-4/02 M:2022: one line per component with its estimate, the value from the certificate or the limits, the distribution (normal from an expanded uncertainty, Type A from a standard deviation and n or from pasted readings, pooled, rectangular, triangular, U-shaped, entered standard uncertainty, custom divisor), the degrees of freedom and the sensitivity coefficient. The tool tabulates the standard uncertainties and contributions, combines them, finds the effective degrees of freedom by Welch–Satterthwaite and the coverage factor from the t distribution (or the dominant-term factors of EA-4/02 S9.14 and S10.13 when one or two rectangular terms dominate), rounds the result as EA-4/02 6.3 asks and writes the statement y ± U (k, p). With tolerance limits it adds the conformity decision of JCGM 106:2012 and ILAC-G8:09/2019: test uncertainty ratio, measurement capability index, the acceptance limits of the chosen decision rule (simple acceptance, guard band w = U, w = rU, a guard band from a target false-accept or false-reject probability, non-binary statements), the conformance probability and the specific risks at the measured value, global risks with a process prior, and the uncertainty the budget would need for a 4:1 ratio. The default is EA-4/02 example S2, the calibration of a 10 kg weight (u = 29.3 mg, U = 59 mg) checked against the OIML R 111 class M1 limits. For an uncertainty propagated through a formula use the Uncertainty Propagation tool.
1 worked example reproduced, listed on the tool page.
Other Quality sections
Control · Design of Experiments · Improvement · Inspection and Metrology