**AIC, AICc, QAIC, and AICc**

The number of parameters in the model is *K*. The AIC depends on the number of parameters as

AIC = -2log Likelihood + 2*K*

and as does the QAIC (quasi-AIC)

QAIC = -2log Likelihood/c-hat + 2*K*

the AICc:

AICc = -2log Likelihood + 2*K* + 2*K*(*K* + 1)/(*n-ess* – *K* – 1)

and the QAICc:

QAICc = -2log Likelihood/c-hat + 2*K* + 2*K*(*K* + 1)/(*n-ess* – *K* – 1)

where *n-ess* is the effective sample size.

You can change the QAIC and QAICc value by changing c-hat with the Adjustments | c-hat menu options from the Results Browser.

An alternative model selection metric is BIC.