If X and Y are two different variables, then through correlation analysis we can find out their relationship with each other. Two variables can either have a linear relationship (when one variable changes in the same proportion as the other one is changing) or non-linear relationship (when change in one variable is not in proportion to the other variable). A linear relationship can be either positive or negative. Correlation and regression techniques are very useful in risk management, marketing research, problem analysis and decision making.
Correlation and its types
Correlation is basically a measure of the association between the two variables. In correlation analysis variables are not designated as independent or dependent. The two mainly used correlation coefficients are: Pearson’s product-moment correlation coefficient and Spearman’s correlation coefficient. Analysts select Spearman’s technique for ordinal data and Pearson’s technique for analysing the ratio-type or interval data. The value of the correlation coefficient may vary between plus one to minus one. A minus one value of the coefficient represents a perfect negative correlation between the variables. On the other hand, a plus one value represents perfect positive correlation between the variables. Zero value of the coefficient simply means there is no relationship between the variables.
When two variable find negative correlation with each other then the value of one variable decreases with the increase in the value of another variable, and vice-versa. In simple words, variables with negative correlation work opposite to each other.
Similarly, when the two variables find a positive correlation with each other then the value of one variable also decreases with the decrease in the value of the another variable. In simple words, variables with positive correlation always move together.
Simple regression is a technique that is used by analysts to examine the exact relationship between one independent and one dependent variable. After performing regression analysis, the results of the analysis can be used to predict the value of dependent variable when the value of independent variable is known to us. Regression analysis goes beyond correlation analysis because of its prediction capabilities.
People use regression analysis on everyday basis. In business, a man who is dressed properly is assumed to be financially fit and successful. A mother always knows that increased sugar in the diet of her children will provide him more energy. Technique of quantitative regression analysis enhances the accuracy in the predictions by developing a mathematical formula for the purpose.
For example, consider a case where a researcher wants to predict the most suitable dose of a new drug (dependent variable) by using the body weight as independent variable. The objective of conducting the regression analysis is to find a mathematical formula that explains the exact relationship between two variables. Then analysts can use the derived formula to predict the values of the dependent variable (dose of the drug in this case) when only the value of independent variable (human weight in this case) is known. So, a doctor can easily prescribe the suitable dose of the drug based upon the body weight of the patient.