Statistics for Data Science and Analytics

10Regression

With correlation, all we can measure is the relative strength of a linear association and whether it is statistically significant. In this chapter, we move on to regression and quantifying how much an outcome variable changes as you change a predictor variable or variables. After completing this chapter, you should be able to:

• Estimate, by eye, a trend line from a two-dimensional scatterplot
• Explain the derivation of a two-dimensional least-squares regression line
• Interpret the meaning of a regression line
• Use a regression line to predict values
• Interpret a residual plot

With regression, we can model that association in linear form and predict values of given values of . The simple form of a linear regression model is as follows:

We read this as “ equals times , plus a constant .” You will note that this is the equation for a line with slope and intercept . The value ...