To run a Bayesian Loglinear Regression model analysis, from the menus choose: Analyze > Bayesian Statistics > Loglinear ModelsFigure 1. Bayesian Loglinear Regression Models dialog
In the Bayesian Loglinear Regression Models dialog, select Gender
[gender] as the Row variable and then select
Employment Category [jobcat] as the Column
variable.
Select Estimate Bayes Factor as the Bayesian
Analysis. The Estimate Bayes Factor option constitutes a natural
ratio to compare the marginal likelihoods between a null and an alternative hypothesis.
Click Criteria to specify analysis criteria for the Bayesian Log-Linear
Regression model.Figure 2. Bayesian Loglinear Regression Models: Criteria dialog
Select the Ascending format option.
Click Bayes Factor to specify the model that is assumed for the observed
data.Figure 3. Bayesian Loglinear Regression Models: Bayes Factor dialog
Select the Multinomial Model option. When selected, the Multinomial model
is assumed for the observed data
Select the Row Sum option under the Fixed Margins
section to specify the fixed marginal totals for the contingency table.
Select Mixture Dirichlet as the Prior Distribution
method.
Click Print to specify how the contents display in the output tables.Figure 4. Bayesian Loglinear Regression Models: Print dialog
Select the Chi-square and Likliehood ratio options
under the Statistics section. The Chi-square option
computes the Pearson Chi-Square statistic, degrees of freedom, and 2-sided asymptotic significance.
The Likliehood ratio option computes the likelihood ratio test statistic,
degrees of freedom, and associated 2-sided asymptotic significance.