Please use this identifier to cite or link to this item:
https://hdl.handle.net/10620/18021| Longitudinal Study: | HILDA | Title: | Income and Wealth Imputation for Waves 1 and 2 | Authors: | Watson, N | Institution: | Melbourne Institute of Applied Economic and Social Research | Publication Date: | Jul-2004 | Publisher: | Melbourne Institute of Applied Economic and Social Research | Pages: | 53 | Keywords: | Nearest neighbour regression missing data income wealth |
Abstract: | All surveys are confronted with the problem of non-response. In an earlier discussion paper, Watson and Wooden (2003) assessed the non-response problem in the HILDA Survey, reviewed various methods for dealing with the missing data, and proposed an imputation strategy. This strategy has been largely adopted for Release 2.0 of the HILDA data. The scope of the imputation has been extended beyond income imputation to include the imputation of the wealth variables that were collected in wave 2. The purpose of this paper is to detail the imputation method used, and discuss the quality of the resulting imputation. | URL: | http://melbourneinstitute.com/downloads/hilda/Bibliography/HILDA_Technical_Papers/htec304.pdf | Keywords: | Surveys and Survey Methodology -- Dataset cleaning and development; Surveys and Survey Methodology | Research collection: | Reports and technical papers |
| Appears in Collections: | Reports |
Show full item record
Items in DSpace are protected by copyright, with all rights reserved, unless otherwise indicated.