Contents
How is QIC used to compare Gee models?
QIC and the related QICu statistic can be used to compare GEE models — that is, for model selection. QICu approximates QIC when the GEE model is correctly specified. QICu, defined as Q+2p, adds a penalty (2p) to the quasilikelihood (Q), where p is the number of parameters in the model.
What is QIC and how is it related to AIC?
The QIC (Quasilikelihood under the Independence model Criterion) statistic proposed by Pan (2001) and further discussed by Hardin and Hilbe (2003) is analogous to the familiar AIC (Akaike’s Information Criterion) statistic used for comparing models fit with likelihood-based methods.
Is there a QIC statistic for Proc Gee?
Beginning in SAS 9.4 TS1M2, QIC is available in PROC GEE. The %QIC macro computes the QIC and QICu statistics proposed by Pan (2001) for GEE (generalized estimating equations) models. These statistics allow comparisons of GEE models (model selection) and selection of a correlation structure.
Is the Akaike information criterion applicable to Gee?
Therefore, Akaike’s information criterion, a widely used method for model selection in GLM, is not applicable to GEE directly. However, Pan (Biometrics 2001; 57: 120–125) proposed a model-selection method for GEE and termed it quasilikelihood under the independence model criterion.
When to use QIC and qicu for model selection?
They provide several guidelines based on the nature of the data for selecting suitable structures that should be applied first. QIC and the related QICu statistic can be used to compare GEE models — that is, for model selection. QICu approximates QIC when the GEE model is correctly specified.
How to select between pre specified models in Gee?
If you want to select amongst pre-specified models, this should work the same with GEE as elsewhere. For example, if you were comparing a nested model to a full model, you could test that. If the models weren’t nested, you could use an informational criterion (such as the QIC) to help adjudicate between them.
When to use% QIC or% qicu macros?
When using QIC or QICu to compare two structures or two models, the model with the smaller statistic is preferred. When using the %QIC macro, the following message will appear but can be ignored: WARNING: Iteration limit exceeded.