Regression and Linear Correlation Analysis

Abstract

Regression and correlation is used in various disciplines such as sociology, biomedicine, engineering, economics, among others. They are closely related, understanding a form of estimation, they are usually used to solve a large number of problems, while the regression is responsible for the relationship between variables, the correlation measures the degree of linear relationship between two or more variables, resulting in the strength and meaning of the relationship.

In other sources consulted, it tells us that the correlation analysis results in a number that summarizes the degree of the correlation between two variables, and the regression analysis gives a mathematical equation that describes that relationship, the data that are necessary come from observations of related variables.

To make a correlation, it is necessary to know what type the variables are, be they quantitative or qualitative, since different methods were used, the advantages of the correlation have their own interpreted language.

In the case of regression, for the result to be correct it is necessary to make an adequate selection of the variables, because if we take variables that do not have any relation to the practice, it will result in illogical, that is, it will not make sense, this is a useful tool for planning.

It is also called regression analysis to any statistical method that establishes an equation allowing estimating the unknown value of a variable from the known value of one or more variables, this term was used for the first time in 1877 by the English statistician Francis Galton, based on a study he did, showing that the height of the tall children tended to recede, or return. It is always developed in an estimation equation, relating the unknown variables with the known variables.

Almost always in the correlation analysis with the regression analysis are used together to measure the effectiveness that the regression line explains the variation of the dependent variable, Y