assignment7

0 40 0 50 0 60 0 7 0 0 1 x 2 x2 40 0 50 0 60

Info iconThis preview shows page 1. Sign up to view the full content.

View Full Document Right Arrow Icon
This is the end of the preview. Sign up to access the rest of the document.

Unformatted text preview: 25 0.50 Ver sus Fi ts (response is Replicates) 0.50 Residual 0.25 0.00 -0.25 -0.50 5.0 7.5 10.0 Fitted Value 12.5 15.0 Comparing these two new plots, there is much improvements in the sense that the residual VS fitted value plots do not show a growth in variance. Also, the normal probability plot shows a more well fitted data to line, hence randomly distributed data. Thus, the last transformation 1/y seems more appropriate for fitting the current regression model. ������������������ ������� � Response Surface Regression: y versus x1, x2, z The analysis was done using coded units. Estimated Regression Coefficients for y Term Coef SE Coef T P Constant 87.3333 1.681 51.968 0.000 x1 9.8013 1.873 5.232 0.001 x2 2.2894 1.873 1.222 0.256 z -6.1250 x1*x1 -13.8333 3.361 -4.116 0.003 x2*x2 -21.8333 3.361 -6.496 0.000 1.455 -4.209 0.003 z*z 0.1517 2.116 0.072 0.945 x1*x2 8.1317 4.116 1.975 0.084 x1*z -4.4147 2.448 -1.804 0.109 x2*z -7.7783 2.448 -3.178 0.01...
View Full Document

This document was uploaded on 03/20/2014 for the course EECS 6.780J at MIT.

Ask a homework question - tutors are online