Bibliography

(eng): The paper demonstrates the fact that applying several estimators (especially with a high breakdown point) of regression coeffs we can obtain rather different result and we are forced to decide which of them is the most "plausible".