Section 8.12 Checklist
When running linear regressions, have you:
-
[ ] Loaded and inspected your data (skim, summary, str)
-
[ ] Cleaned and renamed your variables as needed
-
[ ] Graphed your variables with scatterplots and trend lines
-
[ ] Checked that the relationship looks linear
-
[ ] Calculated correlations before fitting your model
-
[ ] Built your model using lm(y ~ x)
-
[ ] Examined coefficients (intercept and slope) and their significance
-
[ ] Computed predicted values and residuals
-
[ ] Graphed residuals to ensure randomness (no U-shape or funnel)
-
[ ] Tested for homoscedasticity using bptest() (Breusch–Pagan test)
-
[ ] (Optional) Checked residual normality with a histogram or QQ plot
-
[ ] Interpreted R² and Adjusted R² to describe model performance
-
[ ] Added additional predictors for multiple regression if needed
-
[ ] Compared models using AIC or stepwise regression
-
[ ] Selected the most parsimonious model (best fit with the least complexity)
-
[ ] Clearly interpreted what the slope and intercept mean in real-world terms
-
[ ] Reported results visually (scatterplot, line of best fit, residual plot)
