Micromeritics Elzone II 5390 Operator's Manual page 334

Appendix B
SPC Report Variables
Regression Chart
The line of best for the Regression Chart is calculated by the usual least squares method
there is only a single point, or all N points have the same x-value, there can be no line of best
fit in the standard form.
XAve
YAve
Slope
YIntercept
The coefficient of correlation for this line also is calculated in the usual way

x

y
Cov (x,y)
1. BASIC Scientific Subroutines Vol. 11, Ruckdeschel, F. R., BYTE Publications, McGraw Hill, p 16.
2. Mathematical Handbook for Scientists and Engineers, Korn, G.A. and Korn, T.M., McGraw Hill, Sec. 18.4
(1968).
B-8

x
i
-------- -
=
N

y
i
-------- -
=
N

  y 
x – XAve
i i
------------------------------------------------------------- -
=

 XAve 
x –
i
= YAve Slope XAve –

 XAve  2
x –
i
= --------------------------------------
N

2
 YAve 
y –
i
= --------------------------------------
N

 XAve  y
x –
i
------------------------------------------------------------------- -
=
N

YAve
2

 YAve 
–
i
Elzone II 5390 Confirm
2
.
539-42822-01 (Rev A) - Nov 2017
1
. If