Table 5.

Predictions with the two methods: classical calibration and inverse regression, and corresponding radius of the prediction intervals and errors, for data in table 4. In italics the maximum R² and the minimum standard error s.e.

		predictions ${\hat{c}}_i$		prediction interval radius		errors $c_i-{\hat{c}}_i$
Ai	ci	classical $\frac{A_i-{\beta }_0}{{\beta }_1}$	inverse α0 + α1 Ai	classical (a)	inverse (b)	classical ei	inverse ɛi
⋮	⋮	⋮	⋮	⋮	⋮	⋮	⋮
0.181	4	3.62591020	3.7126938	2.582782	2.576339	0.374089799	0.28730621
0.179	4	3.56942588	3.6567619	2.582810	2.576365	0.430574117	0.34323815
0.180	4	3.59766804	3.6847278	2.582796	2.576352	0.402331958	0.31527218
0.209	5	4.41669065	4.4957409	2.582411	2.575986	0.583309349	0.50425915
0.208	5	4.38844849	4.4677749	2.582423	2.575998	0.611551508	0.53222512
⋮	⋮	⋮	⋮	⋮	⋮	⋮	⋮
SSE =						187.5209	185.687
MSE = SSE/(n − 2) =						2.75766	2.730692
$se=\sqrt{\mathrm{MSE}}=,$						1.66062	1.65248
R2 = 1 − SSE/SST =						0.990124	0.9902206