BOOK
Identification of Outliers
Identification of Outliers by David M. Hawkins (1980) presents a comprehensive statistical framework for detecting observations that deviate from the expected structure of a dataset. The book develops both theoretical foundations and practical methods for identifying outliers in univariate and multivariate settings. It emphasizes the impact of outliers on statistical inference, including parameter estimation, regression analysis, and model reliability.
Hawkins introduces classical detection techniques alongside robust statistical approaches designed to reduce the influence of extreme or abnormal observations. A significant contribution of the work is its treatment of multivariate outliers, where relationships among variables are considered rather than examining each variable separately. The book also discusses the concept of influence, showing how certain data points can disproportionately affect statistical conclusions.
No other version available