Summary Results for Random Effects in GLM

The Random effects results are available on the GLM Results - Summary tab.

Element Name Description
Random effects The options under Random effects are used to display the results related to the analysis of the random effects in the model. For a detailed discussion of random effects and the estimation of variance components, refer to the Introductory Overview for the Variance Components and Mixed Model ANOVA/ANCOVA module. Note that the GLM module will only use the ANOVA (expected mean squares) method for estimating variance components; use the Variance Components and Mixed Model ANOVA/ANCOVA module to compute maximum likelihood estimates.
Var. comps. Click the Var. comps (Variance components) button to display a spreadsheet with the ANOVA estimates for the variance components.
Expected MSs Click the Expected MSs button to display a spreadsheet with the elements of the Expected mean squares coefficient matrix, which are used to estimate the variation for each effect in the model.
Bar plot Click the Bar plot button to produce summary plots for the ANOVA estimates of the variance components. Clicking on this button will display a stacked bar plot of the estimated variance components, showing the magnitude of each variance component. If more than one dependent variable has been selected to be analyzed, a compound graph including a stacked bar plot of the variance components for each dependent variable will be produced. If the Plot relative variances check box is selected (see below), relative percentages of the nonzero variance components will be plotted.
Denom. synth. Click the Denom.synth. button to display a spreadsheet with the elements of the denominator synthesis coefficient matrix, which are used to estimate the error term for testing the significance of each random effect in the model.
Pie chart Click the Pie chart button to produce summary plots for the ANOVA estimates of the variance components. Clicking this button will display a pie chart of the estimated variance components, showing the magnitude of each variance component. If more than one dependent variable has been selected to be analyzed, multiple pie charts, one for each dependent variable, will be produced. If the Plot relative variances check box is selected (see below), relative percentages of the nonzero variance components will be charted.
Plot relative variances (% of total) Select the Plot relative variances check box to plot estimated variance components in terms of percentages of total variance when Bar plots or Pie charts are requested (see above). Note that estimated population intraclass correlation coefficients are displayed on bar plots and pie charts when the Plot relative variances check box is selected. These relative variances can be interpreted as zero-order intraclass correlations when there is only one random factor in the analysis. If there is more than one random effect in the analysis and the random effects are correlated, the relative variances should be interpreted as partial intraclass correlations.