STUDYING THE TOTAL UNDER MISSINGNESS BY GUESSING THE VALUE OF A SUPERPOPULATION MODEL FOR IMPUTATION

We propose to use a superpopulation simple regression model. The regression coefficient is predicted by the researcher. The missing values of the variable of interest are predicted using the model. The behavior of the proposed predictor is evaluated by determining its the design expectations of the...

Full description

Saved in:
Bibliographic Details
Published in:Investigación operacional 2020-12, Vol.41 (7), p.979
Main Authors: Bouza, Carlos, N, Viada, Carmen, Vishwakarma, Gajendra K
Format: Article
Language:English
Online Access:Get full text
Tags: Add Tag
No Tags, Be the first to tag this record!
Description
Summary:We propose to use a superpopulation simple regression model. The regression coefficient is predicted by the researcher. The missing values of the variable of interest are predicted using the model. The behavior of the proposed predictor is evaluated by determining its the design expectations of the model bias and variance. A numerical study is developed using real life data. KEYWORDS: superpopulation, imputation, predictor, missing data, assertiveness. Covidl9, Body Index Mass. MSC: 62D05 Proponemos el uso de un modelo superpoblacional del tipo simple regresión. En la regresión el coeficiente es una predicción del investigador. Los valores de la variable de interés cuando se pierde la información son predichos usando el modelo. El compartimento del propuesto predictor es evaluado al determinar su esperanza bajo el diseño así como los sesgos y varianza. Un estudio numérico es desarrollado usando datos de la vida real. PALABRAS CLAVE: superpoblación, imputación, predictor, data faltante, asertividad, Covidl9, Indice de Masa Corporal.
ISSN:0257-4306