On the Comparison of Methods of Estimating Missing Values in Rectangular Lattice Designs

Abimibola V. Oladugba, Emmanuel O. Ossai, Tobias E. Ugah
2018 Journal of Mathematics and Statistics  
Missing values occur in almost all research which lead to ambiguity in data analysis. It becomes necessary that appropriate consideration is made in order to provide an efficient and valid analysis. Researchers have developed and compared a variety of methods of estimating missing values in experimental designs; however, no research work has derived and compared methods of estimating missing values, particularly for rectangular lattice designs. In this study, the Least Square Method (LSM) and
more » ... Method (LSM) and the Analysis of Covariance (ANCOVA) method for estimating missing value in rectangular lattice designs, with and without repetitions, were derived and compared based on four (4) statistical criteria: estimated values, standard errors, p-values and coefficients of determination respectively. Results from the comparison between the derived LSM and the ANCOVA methods showed that the estimates of the LSM appeared more approximate and better than the ANCOVA method in terms of their estimated values, standard errors, p-values and coefficients of determination.
doi:10.3844/jmssp.2018.201.208 fatcat:xd7ckgjgt5dnhn2oupgyiaamw4