A Comparative Analysis of Wheat Yield Using the Range-Control Analysis and the Latin- Square Design

Abstract

Having looked at the use of Latin-Square Design in analyzing agricultural data, it is viewed that a relatively simple alternative can come from statistical quality control. The data set considered in this paper is for wheat yield resulting from a Latin-Square Design. Statistical quality control is basically meant for analyzing data on manufactured products. The data- presentation format for both wheat yield and manufacture data can be the same if the researcher so chooses. One of the possibilities is the data- presentation format for range- control analysis [where the sample size ,n, can be made equal to the  number of samples, m]  and that of the Latin -Square Design where the number of rows is equal to the number of columns is equal to the number of treatments. It is possible to make all these variables equal  for the two methods, and this is what has been done in this paper. The other imposed condition is that if the sample means are significantly different, then, they  are deemed collectively effective. Range-control analysis is a relatively simple method among the methods used in statistical quality   control to determine whether or not the manufactured items meet a pre-manufacture set standard.  For reasons of simplicity, this quality control method has been proposed as a possible alternative to the  Latin-Square Design. The results of the tests conducted using the range- control analysis and the analysis of variance for the Latin –Square design lead to the same statistical conclusions: the effect of rows and columns on the wheat yield is not significant, but the effect of the treatments significantly  influenced the wheat yield.   Hence, it is concluded that the two methods used in the analysis are good alternatives.