PropertyValue
?:abstract
  • Unlike for cross-sectional data, there is only little research on income imputation for long-running panel surveys. In this contribution, we test different longitudinal imputation methods: Little and Su (LS) method, iterative regression with lagged income, and a new imputation method (“mean and within” imputation). The LS method is a univariate approach based on individual mean income over time and is the current best practice for long-running panel data.Iterative regression, which is also frequently used for longitudinal data, has the advantage of using additional information in the wave with missing data. The mean and within approach is based on the individual mean income like the LS method, and adds a component for within variation using iterative regression similar to iterative regression. We evaluate the different imputation methods including complete case analysis using employment income from the Swiss Household Panel from 2000–2021. The nonresponse mechanisms used for the evaluation is based on an external data source containing both registry information and survey questions on income, allowing to detect a not missing at random mechanism. We use performance criteria proposed in previous evaluations of longitudinal imputation methods and add the performance in an application example of regression analysis. Our results confirm the good performance of LS for cross-sectional analysis, but more biased estimates for longitudinal analysis, such as income mobility. The mean and within approach performed best for longitudinal criteria. For multivariate regression, imputation does not improve the estimates. (xsd:string)
?:author
?:comment
  • (SILC) (xsd:string)
?:dataSource
  • EU-SILC-Bibliography (xsd:string)
?:dateModified
  • 2023 (xsd:gyear)
?:datePublished
  • 2023 (xsd:gyear)
?:doi
  • 10.18148/srm/2023.v17i2.7949 ()
?:duplicate
?:fromPage
  • 159 (xsd:string)
is ?:hasPart of
?:inLanguage
  • english (xsd:string)
?:isPartOf
?:issueNumber
  • 2 (xsd:string)
is ?:mainEntity of
?:name
  • Income Imputation in Longitudinal Surveys: A Within-Individual Panel-Regression Approach (xsd:string)
?:publicationType
  • article (xsd:string)
?:reference
?:sourceInfo
  • Bibsonomy (xsd:string)
  • In Survey Research Methods, 17(2), 159-175, 2023 (xsd:string)
?:studyGroup
  • European Union Statistics on Income and Living Conditions (EU-SILC) (xsd:string)
?:tags
  • 2023 (xsd:string)
  • CCBY (xsd:string)
  • FDZ_GML (xsd:string)
  • OA_SSOAR (xsd:string)
  • OAproved (xsd:string)
  • SCOPUSindexed (xsd:string)
  • SILC (xsd:string)
  • SILC_input2023 (xsd:string)
  • SILC_pro (xsd:string)
  • SSCIindexed (xsd:string)
  • article (xsd:string)
  • english (xsd:string)
  • indexproved (xsd:string)
  • jak (xsd:string)
  • review_proved (xsd:string)
  • reviewed (xsd:string)
  • transfer23 (xsd:string)
?:toPage
  • 175 (xsd:string)
rdf:type
?:url
?:volumeNumber
  • 17 (xsd:string)