1351.0.55.159 - Research Paper: Predicting Survey Estimates by State Space Models Using Multiple Data Sources, August 2017
ARCHIVED ISSUE Released at 11:30 AM (CANBERRA TIME) 10/08/2017 First Issue
Page tools: | ||||||||
|
In this research, time series models using multiple data series (SUTSE models) are developed and tested. These models have a number of applications, including as part of a toolkit for identifying and measuring the impacts of statistical process change. Impacts can be assessed via intervention analysis, taking advantage of the cross-correlations and leading properties between the target survey variable and the related series.
Document Selection These documents will be presented in a new window.
|