Which criterion is known as Hurvich and Tsai's criterion for small samples?

Which criterion is known as Hurvich and Tsai's criterion for small samples?

• A. AIC
• B. AICc
• C. BIC
• D. AICC

Answer: (D)

When comparing models with information criteria, small samples bias the standard AIC toward selecting more complex models. Hurvich and Tsai introduced a correction to address this, giving the small-sample corrected AIC. This corrected form adds a penalty term that depends on the number of parameters and the sample size: AICc = AIC + [2k(k+1)]/(n − k − 1). The extra term makes the criterion more conservative when n is small, and as the sample size grows, AICc converges to AIC. In practice, this small-sample criterion is often labeled AICc or AICC. So the Hurvich and Tsai criterion for small samples is the corrected AIC (AICc/AICC).

When comparing models with information criteria, small samples bias the standard AIC toward selecting more complex models. Hurvich and Tsai introduced a correction to address this, giving the small-sample corrected AIC. This corrected form adds a penalty term that depends on the number of parameters and the sample size: AICc = AIC + [2k(k+1)]/(n − k − 1). The extra term makes the criterion more conservative when n is small, and as the sample size grows, AICc converges to AIC. In practice, this small-sample criterion is often labeled AICc or AICC. So the Hurvich and Tsai criterion for small samples is the corrected AIC (AICc/AICC).