Page 35 - 6686

-3
b 1 0,096 > 9,97∙10
-3
b 2 0,242>9,97∙10
All coefficients of equation of regression are significant.

4.5 Evaluation of correctness of the model

The third stage of the analysis of equation is examining
its correctnes.
Suitability of linear equation of regression for finding
optimum area (a set of favorable conditions) is tested by
2
comparing of two variances: correctness S ad and error of
2
experiment S о .
Variance of correctness shows the scattering of average
y
experimental data of variable of condition u relatively those
~
y
values of variable of condition u , which are calculated by
equation of regression:
N
~  b  x , (4.18)
y
 i
u
u 1  iu
2 m N  2
S    y (  y ) (4.19)
ad N  l u 1 u u
where l — number of parts in equation of regression,
remained after examinig their significance.
Accuracy of the model is tested using Fisher’s criterion.
2
S
F  ад , (4.20)
p 2
S
o ,
F  F (4.21)
p T

35