What is the difference between correlation and regression?
Both tell you something about the relationship between variables, but there are subtle differences between the two (see explanation).
Correlation calculates the degree to which two variables are associated to each other. It gives you an answer to, "How well are these two variables related to one another?."
A correlation coefficient ranges from -1 to 1. A correlation coefficient of zero indicates that the two variables are not related in any way, a negative value indicates a negative relationship and a positive value a positive relationship. When measuring for correlation, you would sample randomly the independent and the dependent variables from a population. Correlation makes no assumptions about the relationship between variables. Testing for correlation is essentially testing that your variables are independent.
With regression analysis, one can determine the relationship between a dependent and independent variable using a statistical model. Regression analysis determines the effect of one variable on another. You select values for the independent variable in regression analysis.
The result is a regression equation, which gives you a slope and an intercept and is the average relationship between variables. Regression analysis can be used to predict the dependent variable in a new population or sample. Regression assumes that the dependent variable depends on the independent variable. Regression can also examine multiple independent variables at the same time.
To quickly summarize:
- Strength of association between variables
- Two variables
- Variables assumed to be random
- Assign one variable to be the dependent variable
- Can have multiple independent variables
- Predicts one variable based on the other variable(s)
- Result is an equation