reporting multilevel logistic regression results?

[Vian Rajabzadeh]

I have one more question for today - if I am presenting my multilevel logistic regression results, is it normal practice to report the adjusted and unadjusted in a table? As if it were a single level logistic regression model table of results.

Ciao for now

Vian

[Roman Mostazir]

Ideally, you should provide the log likelihood ratio test (LR) of rho reported at the bottom of the model output. If rho is zero, the multilevel model structure is no different than a pooled estimator. Readers will understand the model improvement by the LR statistic. You don't need to provide both model by tables side by side.

[Vian Rajabzadeh]

Thank you Roman