Announcement

Collapse
No announcement yet.
X
  • Filter
  • Time
  • Show
Clear All
new posts

  • Is it appropriate generating missing values (the Gini index) for future years using MI?

    Dear all,

    I use Stata13 and I am working on 'income inequality' in sub-Saharan Africa (SSA), making use of the Gini indices from Frederick Solt's (2009, 2014) SWIID v4.0 and v5.0. My coverage years are from 1980-2015. However, Gini index for SSA countries ended at 2011. The Gini index is my core argument and the dependent variable, thus meaning that if its years ended at 2011, the econometric analysis for the explanatory variables also ends at 2011 even when they have data till 2015. Thus, I am pondering if I will NOT be committing econometric blunder by doing multiple imputations (MI) to generate indices for years 2012 to 2015? Kindly advise if doing MI is in order, and/or if there is a better way of accommodating missing values without loss of explanatory power on the independent variables.

    Ngozi

  • #2
    In his 2007 paper von Hippel argues that you usually gain little from imputing the outcome/response/dependen/left-hand side variable and even risk making things worse. To the best of my knowledge there is little to do with missing ys.

    Best
    Daniel


    von Hippel, Paul T. 2007. REGRESSION WITH MISSING YS: AN IMPROVED STRATEGY FOR ANALYZING MULTIPLY IMPUTED DATA. Sociological Methodology, 37(1), pp. 83-117.
    Last edited by daniel klein; 14 Oct 2016, 06:24.

    Comment

    Working...
    X