General Flashcards
Cooks distance is calculated by….
The fitting of a new linear model while excluding the outlier in question. The distance of which the new model line deviates from where the old model (model with the outlier included) line is the cooks distance of that outlier.
Cooks distance >1 has a
High influence.
A leverage above 0.25 is
A data point with high leverage and removal should be considered
Influence is related to cooks distance and what axis?
X axis. Although is depicted on the residuals vs leverage graph on the y axis. If above 1 data should be re-considered.
R2 (Rsquared) is
the percentage of variability of Y that is due to X.
SSRegression/SSTotal
6 steps of polynomial regression
- Graph the data and determine a relationship
- Fit an appropriate model
- Test assumptions
- Check test the first of the model
- Pick the best fitted model
- Draw conclusions.