Synthetic balanced study

Balanced crossed Gauge R&R with retained interaction. Calculation accepted; no quality or conformance verdict.

Units: mm; study variation multiplier: 6.

ANOVA

source	df	ss	ms	f	p_value	denominator
part	9	408.10968000000014	45.345520000000015	154587.0000000079	4.1150152144783696e-42	interaction
operator	2	0.20735999999999527	0.10367999999999764	353.4545454545553	3.588398202163839e-15	interaction
interaction	18	0.005279999999999733	0.00029333333333331847	0.23466666666666133	0.9989391652289297	repeatability
repeatability	30	0.03749999999999895	0.0012499999999999651	—	—	—
total	59	408.35982000000007	—	—	—	—

Raw estimates retain negative values. Practical estimates truncate elementary components separately before aggregation.

source	raw_variance	practical_variance	stddev	study_variation	percent_contribution	percent_study_variation
repeatability	0.0012499999999999651	0.0012499999999999651	0.03535533905932688	0.21213203435596129	0.016525741508552065	1.2855248542347233
operator	0.005169333333333216	0.005169333333333216	0.07189807600578207	0.43138845603469245	0.06834165315056713	2.614223654367911
interaction	-0.00047833333333332335	0.0	0.0	0.0	0.0	0.0
part	7.557537777777781	7.557537777777781	2.749097629728304	16.494585778369824	99.91513260534087	99.95755729575471
reproducibility	0.004690999999999892	0.005169333333333216	0.07189807600578207	0.43138845603469245	0.06834165315056713	2.614223654367911
gage_rr	0.005940999999999858	0.0064193333333331815	0.0801207422165645	0.48072445329938696	0.0848673946591192	2.9132008969365497
total	7.563478777777781	7.563957111111114	2.75026491653279	16.50158949919674	100.0	100.00000000000001

PASS means the supported calculation and exports passed checks; it is not a measurement-system quality verdict.
Balanced crossed random-effects model with retained operator-by-part interaction. Each operator must measure every part on the same declared trial identifiers.
Independent, normally distributed, constant-variance errors and representative random parts/operators are assumptions; the file cannot establish them.
Trial identifiers identify replicates; they do not model time trends or systematic trial effects. No pooling, imputation, nested/destructive study, tolerance or process-acceptance calculation.
Raw method-of-moments component estimates may be negative. Practical components truncate each of the four elementary estimates to zero, then sum them; these are distinct from raw estimates.
Study variation is six times the practical standard deviation. Percent contribution sums to 100 across elementary components; percent study variation does not.
No confidence intervals, sampling-design certification, calibration, manufacturing approval, clinical application or regulatory conformance claim.
Finite-precision numeric calculation: sum-of-squares dust at or below total SS times 1e-28 is reported as zero. Total SS at or below 1e-24, or residual/interaction MS at or below total SS times 1e-14, is rejected. Small reported p-values can underflow to zero.

Source SHA256: 50093854e5d0b7895912ee856bc62ea2da97b81480b0bcd308a9e38680f8848a

Algorithm: gaugestudy-1.0