Analyze Variability
MLE estimation method - P-Values
|
|
Analyze VariabilityMLE estimation method - P-Values |
|
Minitab provides two methods for fitting your model: least squares regression and maximum likelihood estimation. In many cases, the differences between the results of the two methods are minor. One approach is to use the p-values from the least squares analysis to choose the terms in the model, then use the coefficient estimates from the maximum likelihood analysis to calculate fits.
Example Output |
|
|
Coded Coefficients for Ln(Std)
Ratio Term Effect Effect Coef SE Coef Z-Value P-Value VIF Constant 0.3538 0.0791 4.48 0.000 Material -0.9607 0.3826 -0.4803 0.0791 -6.08 0.000 1.00 InjPress -0.1830 0.8328 -0.0915 0.0791 -1.16 0.247 1.00 InjTemp 0.0566 1.0582 0.0283 0.0791 0.36 0.720 1.00 CoolTemp -0.1204 0.8866 -0.0602 0.0791 -0.76 0.446 1.00 Material*InjPress -0.9927 0.3706 -0.4963 0.0791 -6.28 0.000 1.00 Material*InjTemp 0.1866 1.2051 0.0933 0.0791 1.18 0.238 1.00 Material*CoolTemp 0.0032 1.0032 0.0016 0.0791 0.02 0.984 1.00 InjPress*InjTemp -0.0775 0.9254 -0.0388 0.0791 -0.49 0.624 1.00 InjPress*CoolTemp -0.0800 0.9231 -0.0400 0.0791 -0.51 0.613 1.00 InjTemp*CoolTemp 0.0128 1.0129 0.0064 0.0791 0.08 0.935 1.00 |
Interpretation |
|
In this example, the manufacturer already has the least squares regression results, which showed that material and the material by injection pressure interaction were significant with p-values equal to 0.000. The MLE results confirm the least squares regression results. The factor material and the interaction material by injection pressure are both significant with p-values equal to 0.000.
