Prob > ChiSquare

Degrees of Freedom

Difference: Chi-sq

After Fitted Variances

Equal Variances Initially

Fit

Marginal Variances

Sqrt(Marg Var)

Marg.Variance

L-R ChiSquare

Model

Effect Likelihood-Ratio Tests

Variance

Mean

Newton

StepHalve

-2*LogLikelihood

Confid. Limit per row, showing corresponding values of other parameters at limit

Side

Parameter

Confidence Interval Detail

Upper

Lower

Upper 

% Limit

Lower 

Specify alpha level for confidence interval 

-2*LogLikelihood

Source

Likelihood Ratio Test for Equal Variance

exp(2|Estimate|)

exp(Estimate)

t Ratio

Std Error

Variance Parameter Estimates

LogVariance Fit does not distinguish Random Effects.

No Multiple Nominal Y's allowed.

LogVarStd

LogVarWgt

LogVarResid

No Variance Effects, Wrong Fitting Personality

Do not use for production work.

LoglinearVariance Fitting is an untested experimental feature.