Table 8. Regression models

	Radial direction	Tangential direction
p	0.95	0.95
Gr	2.116 < 2.176	2.033 < 2.176
	No major misses	No major misses
G	0.0468 < 0.0865	0.0444 < 0.0865
The hypothesis of homogeneity of variances	Accepted	Accepted
S2y	8.999·10−3	9.535·10−3
Δbi	28.8·10−3	31.6·10−3
δ	δ r = π ⋅ Q r − 1 = = 2.912 − 0.237 x 1 + + 0.057 x 2 + 0.379 x 3 + + 0.435 x 4 + 0.253 x 5 − − 0.042 x 1 x 4 − 0.031 x 3 x 4 − − 0.033 x 1 x 2 x 5 + 0.040 x 1 x 3 x 4 − − 0.034 x 1 x 2 x 3 x 4	δ τ = π ⋅ Q τ − 1 = 2.559 + 0.051 x 1 + + 0.074 x 2 + 0.304 x 3 + + 0.356 x 4 + 0.131 x 5 + + 0.078 x 1 x 2 + 0.073 x 1 x 5 − − 0.063 x 1 x 2 x 5 + 0.063 x 1 x 3 x 5 − − 0.054 x 1 x 2 x 3 x 4
l	11	11
Fad	F a d = 4.252 ⋅ 10 − 3 8.999 ⋅ 10 − 3 = 0.472 < 2.764	F a d = 7.704 ⋅ 10 − 3 9.535 ⋅ 10 − 3 = 0.808 < 2.764
	The model is adequate	The model is adequate

p: probability, Gr: Grubbs’ criterion, G: Cochran’s criterion, S²_y: the reproducibility variance of the experiment, Δb_i: the significance level, δ: the logarithmic decrement of damping, l: the number of significant coefficients in the regression equation, F_ad: Fisher F criterion.