For inference in complex models, composite likelihood combines genuine likelihoods based on Q3 low-dimensional portions of the data, with weights to be chosen. Optimal weights in composite likelihood may be searched following different routes, leading to a solution only in scalar parametermodels. Here, after briefly reviewing themain approaches, we show how to obtain the first-order optimal weights when using composite likelihood for inference on a scalar parameter in the presence of nuisance parameters. These weights depend on the true parameter value and need to be estimated. Under regularity conditions, the resulting likelihood ratio statistic has the standard asymptotic null distribution and improved local power. Simulation results inmultivariate normal models show that estimation of optimal weights maintains the standard approximate null distribution and produces a visible gain in power with respect to constant weights.

Efficient composite likelihood for a scalar parameter of interest

Luigi Pace
;
2019-01-01

Abstract

For inference in complex models, composite likelihood combines genuine likelihoods based on Q3 low-dimensional portions of the data, with weights to be chosen. Optimal weights in composite likelihood may be searched following different routes, leading to a solution only in scalar parametermodels. Here, after briefly reviewing themain approaches, we show how to obtain the first-order optimal weights when using composite likelihood for inference on a scalar parameter in the presence of nuisance parameters. These weights depend on the true parameter value and need to be estimated. Under regularity conditions, the resulting likelihood ratio statistic has the standard asymptotic null distribution and improved local power. Simulation results inmultivariate normal models show that estimation of optimal weights maintains the standard approximate null distribution and produces a visible gain in power with respect to constant weights.
File in questo prodotto:
File Dimensione Formato  
sta4.222.pdf

non disponibili

Tipologia: Versione Editoriale (PDF)
Licenza: Non pubblico
Dimensione 248.88 kB
Formato Adobe PDF
248.88 kB Adobe PDF   Visualizza/Apri   Richiedi una copia
Efficient composite likelihood for a scalar parameter of interest.pdf

accesso aperto

Tipologia: Documento in Post-print
Licenza: Creative commons
Dimensione 112.83 kB
Formato Adobe PDF
112.83 kB Adobe PDF Visualizza/Apri

I documenti in IRIS sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.

Utilizza questo identificativo per citare o creare un link a questo documento: https://hdl.handle.net/11390/1144459
Citazioni
  • ???jsp.display-item.citation.pmc??? ND
  • Scopus 8
  • ???jsp.display-item.citation.isi??? 8
social impact