Relationship between correlation and regression

Explain the relationship between correlation and regression. What do the two approaches
have in common, and how are they different? In what types of research might one
approach be preferred over the other? Explain and justify your response. Be sure that your
discussion refers to both predictor and criterion variables as well as their role in the
regression equation

Relationship between correlation and regression

Correlation and regression are some of the most common and useful methods of statistics
used by statisticians. Correlation is used to describe the degree of the relationship between two
variables while on the other hand regression is used to describe the relationship between a single
dependent and independent variable. It therefore helps to tell us about the linear association that
exists between two variables.
There are some similarities between these two approaches. One is that regression is
derived from correlation as regression equation is obtained once the correlation value is known
(Jon, 2000). Another similarity is that both correlation and regression analysis deal with
relationships among variables. Furthermore, both analyses are not interpreted as causing cause-
and-effect relationship but rather they indicate how or to what extent variables are associated.

Consequently, correlation and regression have some differences. There is no reference to
causation in correlation and there is no variable considered as independent or dependent. The
two separate variables have specific relationship and co-relate. On the other hand, in regression
one of the variables is considered to be independent and is also called predictor and the other
dependent variable also called criterion variable. Therefore, in regression analysis one variable

causes certain percentage of change on the other variable (Jon, 2000). These two variables are
important in the regression equation because they help to describe the relationship between these
two variables. Results of regression are more generalizable compared to those from correlation.
Correlation is applicable in various research work which includes observational research,
survey research and archival research such as violence and economic. This is because; it helps to
give the degree of relationship or association between variables. Regression analysis is
applicable in quantitative researches as it helps to describe the relationship that exists between
the independent and dependent variables.


Jon, R. (2000). Correlation or regression.