There are two concepts of regression analysis that will help define what regression
analysis is. If you understand these two concepts you will be well on your way to
understanding what regression analysis is all about. You have probably already visited the
Simple Regression Calculation page
on this website that provides a "live" example
of calculating linear regression. The results of this analysis are the two
parameters that we want to discuss here.

On other pages of this website the value and importance of regression was discussed.
The emphasis here is on the interpretation and the validity of the analysis after it
has been preformed. There are many forms of regression including multiple regression,
exponential regression, logarithmic and even sine wave regression. The only
regression that will be discussed here is linear regression. It is the simplest,
but all the same principles apply to the other forms. The two concepts of
regression analysis that will be discussed are:

The equation calculated from the data.

The correlation between the data and the calculated line.

The Equation

The first important concept that you must have when you view the results of a
linear regression calculation is, you are given the parameters for the equation of
a line.

M is the slope and b is the y intercept. Looking at the graph you can see
what the calculated line looks like.

The Correlation

The second important concept that you must have when you view the results of
a linear regression calculation is the correlation. The correlation is the r
value. It can have a value between -1 and 1. How can do you know if you can
rely on equation described above? The correlation helps you determine what
level of confidence you can have. The closer to 1 that r is, the greater
confidence you have. For a more detailed look at correlation go to
Correlation Defined.

Summary

The two concepts of regression analysis that are most important are:

The equation that it gives you.

The correlation coefficient.

With these two concepts you can understand what the result of the calculation is and
whether it can be relied on as a predicter.