Scatter Plots And Looking at scatter plots Or Bivariate Data
What are we doing? Finding the “Least Squares Line” What do we use to find the variables “a” and “b”? Why do we do it? So we can predict an outcome.
manateespowerboats We need to use the data to calculate a Best fit line. Best fit, is also referred to as least squares line. Enter this data into your TI-Nspire CAS. Open a New Document, add lists and spreadsheets.
Create a scatterplot of the data by: CTRL I 5 select “manatees” as the y-variable Select “powerboats” as the x-variable. We can have the calculator assign us a least squares line by completing the following procedure: Menu, 4, 6, 2 -or- Menu, Analyze, Regression, y=a+bx
Even though we have a Line of Best Fit, this does not mean that this is the best model for our data. One last check is to look at the Residuals: Menu, 4, 7, 2 -or- Menu, Analyze, Residuals, Show Residual Plot T here should be an equation near the straight line on your scatterplot. Check to make sure your equation matches the one to the right. This is your Line of Best Fit or Regression Line.
Your plot should now be divided. Scatterplot with Regression Line on top, Residual Plot with a horizontal line on the bottom of the screen. Residual Plot: Tells you if your model (equation for predicting) is actually appropriate. Ideally, you would like to see the dots completely scattered – no patterns or curves appearing, and you don’t want to see any dots very far away from the horizontal line. If your residual can meet these 2 conditions, YOU HAVE A GREAT MODEL!!!!!!
What can the residual tell you? If you your residual is negative, your predicted value is TOO HIGH.