Equation of line of best fit and regression?

Question

How do i determine the line of best fit and regression coefficient for a set of data (Ex: Number of weeks emploed vs. questions asked.)?

Answer 1

Refer to a statistics text. There are different ways of mathematically defining the "best fit" and, of course, you can try to fit the data to a linear, quadratic, or cubic, etc function. The regression coefficient (Rsq) and the adjusted coef (RsqA) are measures of how well the data fit the equation. It is best if you decide up front how close the equation should fit to the data for you to decide it is a "fit". There are some statistical assumptions involved and, if you are doing this analysis rigorously, you should identify what those assumptions are and check them.



Statistical software packages such as Minitab do all the math for you very quickly and easily. However, if you don't have a reasonably good idea of what it all means and what the assumptions are, you may not know as much as you think you know.



In basic terms, a best fit line is a mathematic model. Someone wise once said that all models are wrong, some models are useful.