Fixed Effects Tests
In the Generalized Linear Mixed Model report, the Fixed Effect Tests section contains a table of significance tests for each fixed effect in the model. The test for a given effect tests that the null hypothesis that all parameters associated with that effect are zero. An effect, such as a single continuous explanatory variable, might have only one parameter. In this case, the test is equivalent to the t test for that term in the Fixed Effects Parameter Estimates report.
The Fixed Effects Tests report contains the following columns:
Source
The fixed effects in the model.
Nparm
The number of parameters that are associated with the effect. A continuous effect has one parameter. The number of parameters for a nominal or ordinal effect is one less than its number of levels. The number of parameters for a crossed effect is the product of the number of parameters for each individual effect.
DFNum
The numerator degrees of freedom for the effect test.
DFDen
The denominator degrees of freedom for the effect test, or the degrees of freedom for error. DFDen is calculated using the Kenward-Roger first order approximation.
F Ratio or ChiSquare
The computed F ratio or chi-square statistic for testing that the effect is zero. If no random effects are present in the model and the specified Distribution has only one parameter, the test is a chi-square test. Otherwise, the test is an F test.
Prob > F or Prob>ChiSq
The p-value for the effect test.