An adjusted quasi-profile loglikelihood for a scalar parameter of interest is discussed. The corresponding quasi-profile score has bias which is asymptotically negligible. Because of this, it is likely to provide accurate inferences with small samples or many nuisance parameters. An application to robust M-estimation is presented, which gives rise to a robust adjusted quasi-profile loglikelihood.
Adjusted quasi-profile likelihoods and robustness / R., Bellio; Brazzale, Alessandra Rosalba; L., Ventura. - STAMPA. - (2001), pp. 413-416. ((Intervento presentato al convegno 16th International Workshop on Statistical Modelling tenutosi a Odense nel 2-6 luglio 2001.