Seminar Notice
                Statistical Laboratory
                 Iowa State University

DATE AND TIME:     Thursday, November 6, 1997, 1:10 p.m.

PLACE:             171 Durham

SPEAKER:           Marek Brabec
                   Department of Statistics
                   Iowa State University
                   
TITLE:             Simple Adjustment Procedure for Asymptotic Bias
                   Removal and Its Applications

                       ABSTRACT

We develop a two-step estimation technique termed Adjusted Quasi
Maximum Likelihood Estimation, (AQMLE). It starts with a quasi
maximum likelihood estimate (QMLE), obtained by maximization of an
incorrect likelihood (quasilikelihood). The quasilikelihood is often
a simplified approximation to the correct likelihood, used to avoid
excessive numerical computations required to maximize the true
likelihood. On the second step, the QMLE is adjusted to remove
asymptotic bias. The adjustment is defined implicitly through the
expectation of the quasiscore under the correct model. When the QMLE
is easy to obtain, the AQMLE has an advantage of rather stable
computational requirements even for large n. Hence it provides a
practical alternative to the MLE, when evaluation of the correct
likelihood is complicated or when complete data are not accessible.
The AQMLE is consistent and asymptotically normal, allowing for
construction of approximate tests and confinfidence intervals.
Finite sample behavior is explored in a censored lognormal-normal
convolution model, describing a practical situation encountered when
the concentration of a chemical substance in a particular material
is measured with error and data are censored as the result of a
detection limit. In a simulation study, the AQMLE is compared with
the rarely used MLE (requiring separate numerical integration for
each observation) and a commonly used but asymptotically biased QMLE.