|Learning by Simulations has been developed by Hans Lohninger to support both teachers and students in the process of knowledge transfer and acquisition . Click here for more information.|
|Home Statistics Leverage Effect|
|See also: Linear Regression, Regression of Weakly Correlated Data|
Share this Page:
The term "leverage" is commonly used for an undesirable effect which is experienced with regression analysis (as well as with other methods). It basically means that a single data point which is located well outside the bulk of the data (an "outlier") has an overproportional effect on the resulting regression curve. Depending on the number of the samples and the distance of the outlier from the rest of the data, this effect may completely corrupt a regression model which would be quite good in the absence of an outlier.
At the left you see a typical example of the leverage effect. The blue data point at the lower right of the diagram leads to a tilting of the regression line and a considerable reduction of the goodness of fit (r2 = 0.77 (including the outlier) vs. 0.98 (without the outlier))