Ja. Osborne et Ta. Severini, Inference for exponential order statistic models based on an integrated likelihood function, J AM STAT A, 95(452), 2000, pp. 1220-1228
Methods of statistical inference are developed for the exponential order st
atistic (EOS) model, where only a subset of order statistics from a collect
ion of N lid exponential random detection times is observable. When the rat
e parameter for detections is unknown, the maximum likelihood estimator (ML
E) of the unknown integer parameter N can be infinite with substantial prob
ability. inference for N is developed using a pseudolikelihood function obt
ained by integrating out the rate parameter. The estimator that maximizes t
his function, called the integrated likelihood estimator (ILE), is shown to
be finite and to have better sampling properties than the MLE. Parameter-b
ased asymptotics are developed for the case where N is large. Application o
f the methodology is illustrated using two datasets.