outlier in robust regressiom

#1
hi, I dont understand with this statment.

"robust regression can be used to detect outliers and to provide resistant
(stable) results in the presence of outliers"

what the meaning of " to provide resistant (stable) result inthe presence of outliers"?
what the meaning of "resistant" in that statment?

thankyou.
 

hlsmith

Omega Contributor
#2
Do you know what an outlier is in regression?


-An extreme value that is not near the predicted regression line or general clustering of the observations. Could be measured by the number of standard deviations away from the mean error. These values can leverage/influence the estimate of the fitted line compared to the model's fit if the value was removed. Robust regression functions to have resistant results, meaning the outlier doesn't influence the model as severely as in perhaps ordinary least squares modeling.


May help to Search web for images/picture related to outliers and also robust regression.