Using Eta (η) Correlation Ratio In Analyzing Strongly Nonlinear Relationship Between Two Variables In Practical Researches

Abstract

This paper main aim is Eta (η) correlation ratio in analyzing strongly nonlinear relationship between two Variables in breeding area researches. In researches on a relationship that leads two –Variables correlation to determine correlation coefficient if two Variables would Continual (interval, ratio) and parametric uses Pearson torque correlation and if discrete (ordinal, nominal) or nonparametric uses Spearman Brown. And to determine regression line writes Predicting equation as linear relation and if Variables would be more than two uses multivariable regression. But in many cases relationship between Variables is nonlinear that is useful to analyze data use minimizing practice method. This method could expand in many nonlinear senses. But if relationship between two Variables [both continual] would be strongly nonlinear using minimizing is not suitable. That in this correlation as researcher uses Eta (η) correlation ratio instead of correlation coefficient obtains efficient result to specify meaningful difference. This paper refers to common methods inefficiency and defects with an example then refers to (η) advantages, efficiency and how calculating and how using it in practical researches.