Methods for Utilising Partially Missing Data in Two-Phase Sampling

Show simple item record

dc.contributor.advisor Wild, C en
dc.contributor.author Grimson, Fiona en
dc.date.accessioned 2011-02-01T01:04:12Z en
dc.date.issued 2011 en
dc.identifier.uri http://hdl.handle.net/2292/6145 en
dc.description Full text is available to authenticated members of The University of Auckland only. en
dc.description.abstract We consider the analysis of two-phase samples where a large cohort is sampled at the first phase with a smaller subsample taken at the second phase. Some variables are observed at the first phase for the whole cohort and some at the second phase only for the subsample. Methods of utilising the partial observations to improve estimation of regression coefficients are investigated. We investigate methods that involve the use of partial information to estimate auxiliary variables, rather than impute missing data values. We focus on the comparison of three methods which are all of the calibration class of estimators and are asymptotically equivalent. We wish to assess their performance and potential to increase efficiency over standard analysis techniques infinite samples. This is done through simulation studies comparing the performances of these methods in a variety of settings, including both binary and continuous response variables, stratified data and the availability of surrogate information. A substantial finite sample effect was observed. The partial data methods were found to perform well, particularly the iterated methods. An advantage was also found using direct estimation of the optimal auxiliary variables. We also consider methods of estimating the variance of parameter estimates generated by these methods from a single data set. Variance estimation methods are also tested through simulation studies. en
dc.publisher ResearchSpace@Auckland en
dc.relation.ispartof Masters Thesis - University of Auckland en
dc.relation.isreferencedby UoA99216072814002091 en
dc.rights Restricted Item. Available to authenticated members of The University of Auckland. en
dc.rights.uri https://researchspace.auckland.ac.nz/docs/uoa-docs/rights.htm en
dc.title Methods for Utilising Partially Missing Data in Two-Phase Sampling en
dc.type Thesis en
thesis.degree.discipline Statistics en
thesis.degree.grantor The University of Auckland en
thesis.degree.level Masters en
dc.rights.holder Copyright: The author en
pubs.elements-id 202308 en
pubs.record-created-at-source-date 2011-02-01 en


Files in this item

Find Full text

This item appears in the following Collection(s)

Show simple item record

Share

Search ResearchSpace


Browse

Statistics