ANALISIS KOVARIAN PADA RANCANGAN BUJURSANGKAR GRAECO LATIN

Abstract

Analysis of covariance (ancova) is a technique that combines features of analysis of variance and regression which is used to increase precision (accuracy) of the experiment. Ancova can be used for any experimental design include Graeco Latin square design. Graeco Latin square design is a combination of two orthogonal Latin square, two Latin square are orthogonal if when they are combined, the same pair of symbols occurs no more than once in the composite square. Application ancova on Graeco Latin square design in the field of agriculture is given to observe the effect of different fertilizer dose towards outcome of corn production. In this experiment there are three blocking factors (soil pH, soil slopes, and corn varieties) and two variable concomitant (quantity of corn plant and quantity of baby corn). The result shows that both of concomitant variables effect the respons, that’s mean using these two concomitant variable on this ancova method is appropriate. More than that, ancova give result way much better than anova, shows from the result coefficient of variation ancova less than anova, means that there is an increase precision (accuracy) of the experiment. So, concomitant variable can’t be ignored from the experiment. Keywords: Analysis of Covariance (ancova), Graeco Latin square design, Analysis of Variance (anova)