Data imputation in trials with genotype
by environment interaction: An application
on cotton data

Sergio ARCINIEGAS-ALARCÓN[1]

Carlos Tadeu dos Santos DIAS[2]

§    ABSTRACT: A common problem in multienvironment trials are the missing genotype-environmental combinations. Recently, Bergamo proposed a distribution-free multiple imputation method in the interaction matrix. The purpose of this paper is to evaluate the new development and compare it with methodologies that have success in the genotype-environmental trials with missing data, like the alternating least squares (ALS) and the robust estimates, using the Additive Main effects and Multiplicative Interaction Models (AMMI). Was made an simulation study based in real data, doing missed random considering different percentages, imputing the observations and comparing the methodologies through three criteria: the square root of the mean predictive difference, the Procrustes statistic and the Spearman's rank correlation coeficient.  Was concluded that the multiple imputation is not better than the imputation based in a additive model without interaction, and the best results for the variance are obtained with robust sub-models.  All the considerated methods in this study have a high correlation between the true and the imputed missing values.

§    KEY WORDS: Missing data; data imputation; AMMI models; genotype-environment interaction.

 



[1] Programa de Pós-Graduação em Estatística e Experimentação Agronômica, Escola Superior de Agricultura “Luiz de Queiroz” – ESALQ, Universidade de São Paulo – USP, Caixa Postal 9, CEP: 13418-900, Piracicaba, SP, Brasil. E‑mail: salarcon@esalq.usp.br

[2] Departamento de Ciências Exatas, Escola Superior de Agricultura “Luiz de Queiroz” – ESALQ, Universidade de São Paulo – USP, Caixa Postal 9, CEP: 13418-900, Piracicaba, SP, Brasil. E‑mail: ctsdias@esalq.usp.br