Deming vs Passing Bablok at 3 x 3 digits precision
Here the result of a successful simulation with 400 replicas per point at 3x3 digits precision. Classical PaBa still shows a pretty high asymmetry with r.s.e of over 4%. PBequi is performing better, but shows some bias compared to classical Deming methods with an r.s.e over 2%. Quite surprisingly PBequi with jackknife CI performs better that the analytical version with approx. 1.36% r.s.e.
| Methods | r. s. vol. ratio | r.s.e% |
|---|---|---|
| Deming - anal. | 0.5075293 | 1.5058635 |
| Deming - jack. | 0.5059993 | 1.1998674 |
| Deming - boot BCa | 0.4998623 | -0.0275387 |
| PBequi - anal. | 0.5105528 | 2.1105656 |
| PBequi methodlarge=F - anal. | 0.5112202 | 2.2440393 |
| PaBa - anal. | 0.5220235 | 4.4047071 |
| PBequi - jack. | 0.5067896 | 1.3579233 |
| MDeming - jack. | 0.5075982 | 1.5196464 |
| MMDeming - jack. | 0.5113902 | 2.2780404 |
| Deming - JE MD 1% | 0.5030620 | 0.6123919 |
Table of r.s.e. at 3x3 digits precision
Below is also reported a plot for the M- and MM-Deming methods. M-Deming is very well performing, in line with other classical Deming regressions with a r.s.e% of approx. 1.52%. MM-Deming is confirming its bias with an r.s.e% of 2.28%.
Worth noting that the Deming JE/MD method has an r.s.e% of only 0.62%. This method seems much more powerful that the traditional ones and also shows a good symmetry.
