1. Simple regression
2. Multiple regression
   1. Considerations variables
      1. Logic
      2. Fit
      3. Parsimony
      4. Stability
   2. Tests for relevance variables
      1. Statistically significance
         1. Overall fit: F test
            1. H0: B1 = B2 = Bn = 0
         2. Individual Beta t test
            1. H0: B1 = 0
            2. H0: B1 >= X
            3. H0: B1 =< X
         3. Partial F test
            1. H0: B1 = B2 = 0 out of a total of Bn
            2. H0: B1 = B2
            3. H0: B1 + B2 = 1
      2. Overall Practicality
         1. R squared: Coefficient of determination
         2. Adjusted R Squared
         3. Standard error of the regression
   3. Considerations for multiple variables
      1. Binary predictor (dummy variable)
      2. Interaction between variables
         1. Test significance of the coefficient of X1*X2
      3. Multicollinearity: indepencany of variables
         1. Test: Variance Inflation Factor (VIF)
3. Conditions / assumptions / points of interest
   1. Residual errors
      1. Errors are normally distributed
         1. Test: Histrogram standardized residual
         2. Test: Normal probability plot of the residuals
         3. Test: Skewness and Kurtoisse
      2. Errors have constant variance: Homoscedastic
         1. Test: Scatterplot residuals (for simple regression)
         2. Test: Plot residuals against predicted Y values (multiple regression)
      3. Errors are independent (non autocorrelated)
         1. Test: graph Residuals
         2. Durbin Watson test
   2. Unusual observations
      1. Leverage
      2. Outliers
   3. Linearity
      1. Test: adding Xi^squared as variable
4. Predictions
   1. Y
      1. Individual prediction
      2. Confidence interval Individual
         1. Simple regression
         2. Multiple regression
      3. Confidence interval Mean
         1. Simple regression
         2. Multiple regression
   2. Bi
      1. Individual prediction
      2. Confidence interval Individual