Online Program

Imputation in High Dimensional Economic Data as Applied to the Agricultural Resource Management Survey (ARMS)
*Michael Robbins, University of Missouri 

Keywords: Missing Data, Imputation, ARMS, Markov Chain Monte Carlo, Gaussian Copula.

This talk concerns imputation imputation in the USDA's ARMS data, which is a high dimensional economic dataset. A robust joint model for these data which requires that variables are transformed using a suitable class of marginal densities is developed. Thus, the variables may be linked through a Gaussian copula, which enables construction of the joint model via a sequence of conditional linear models. The criteria used to select the predictors for each conditional model is discussed. For the purpose of developing an imputation method that is conducive to these model assumptions, we propose a regression-based technique that allows for flexibility in the selection of conditional models while providing a valid joint distribution. In this procedure, parameter estimates and imputations are obtained using a Markov chain Monte Carlo sampling method. Next, by repeatedly poking holes in observed data using various mechanisms via simulation, the performance of the proposed technique is analyzed. Finally, the proposed method is applied to the full ARMS data, and brief data analysis that serves to gauge the appropriateness of the resulting imputations is presented.