Vol. 16 No. 3
Application of ANCOVA Model for estimation of missing observation(s) and adjusted treatment means in split plot layout
Author(s): M. SAHA, J. S. BASAK AND A. MAJUMDER
Abstract: In agriculture, split plot experiments are becoming very popular among the researchers to quantify the responses of many factors through a single experiment. But sometimes, data loss in an experiment leads to major problem for analysis of experiment. Proper estimation of these missing observations from the data set leads to restore the orthogonal property of the designed experiments. The present study aims to apply the ANCOVA model in split plot layout with missing observation(s) as well as to compare different treatment mean values on the basis of adjusted mean values using regression coefficients of concomitant variables on response variables. The application of the above model for estimation of missing values (single and two) and analysis methods are showing very optimistic results in the field experiments on four yield parameters of maize during 2017-18. The absolute difference between the estimated missing value and the actual value is negligible in most of the cases. Moreover, the analysis by ANCOVA model with one and two missing observations shows that the adjusted treatment means along with their first order interactions (Federer, 1975) are same with the original data set which contains no such missing observations.
Download Full Article: Click Here