Procedure Overview

1. Given paired numerical variables, decide which variable is independent (x) and which is dependent (y). 
2. Draw a scatterplot of the data.
3. Examine the scatterplot to determine the basic shape of the curve.  For example, the shape might be one of the following.  (Click the links below if you need a review of the basic shapes and then click the "back" button on the browser instead of the buttons below the graphs.)

   • Linear
   • Quadratic
   • Cubic
   • Quartic
   • Exponential
   • Logarithmic
   • Sinusoidal

4. If there appears to be a linear association between the two variables, then find the sample correlation coefficient of the variables.  (In the Correlation Tutorial, for the temperature/water example we found the linear correlation to be r=0.96.) 
5. If the strength of the association is strong enough, then use the procedures described in this tutorial to find the line of best fit in the sense of least squares.
   

The Temperature/Water example will be used to demonstrate the procedure to calculate the equation of the Least Squares Regression Line by-hand, using MS Excel, and using Minitab.