Leverage, Studentized Residual, and Cook's Distance

In this activity, you can explore how leverage, studentized residual, and Cook's distance reflect the extremeness and influence of a case. Case A (the red dot) can be moved. Try to move it and see how leverage, studentized residual, and Cook's distance change. You can also see how these changes are related to the differences between two models: One with all cases, and one with Case A excluded (e.g, regression coefficients, R-squares). On the right, the studentized residuals are plotted against centered leverage values. Because Cook's distance can be computed from studentized residual and leverage, you can also use the slide to set a Cook's distance and the find the two lines with Cook's distance equal to this value. Cases above the upper line or below the lower line have Cook's distance larger than the value you set. Sample size also affects how influential a case is. Therefore, you can change the sample size, and see how the regression line is affected by Case A. Last, you can find two grey crosses. You can move these two crosses to change the population model used to generate the data.
Note: I enabled the right click menu. In each graph, you can right click to get the context menu, and then choose "Zoom to fit" or other zoom options. I will update other activities soon.