Informativity and robustness
This paper investigates robustness problems involving non response or other nuisance parameters. The minimal sufficient parameter is introduced and shown to completely capture the information provided by data. The concept of informativity is then introduced that serves to determine, in some sense, whether statistical models provide any information about inferential quantities of interest. These concepts are shown to lead to useful robustness results involving inferences under perfect...[Show more]
|Collections||ANU Research Publications|
|05-09.pdf||231.96 kB||Adobe PDF|
Items in Open Research are protected by copyright, with all rights reserved, unless otherwise indicated.