×

zbMATH — the first resource for mathematics

Outliers in statistical data. 2nd ed. (English) Zbl 0638.62002
Wiley Series in Probability and Mathematical Statistics. Applied Probability and Statistics. Chichester etc.: John Wiley & Sons. XIV, 463 p. (TUB, FB Stat.: B 86271) (1984).
[For a review of the first edition from 1978 see Zbl 0377.62001.]
The modifications in the revised edition are of three types. Firstly, there are areas of enquiry which represent new topics of outlier study or which now need to be described within the context of outlier methodology. These are reflected in the new chapters on outliers in directional data, and in time series, respectively.
Secondly, there are the many contributions which refine, reassess, or extend our knowledge on specific aspects of outlier investigation. Such developments are incorporated by means of substantial expansion, and some judicious pruning, of the discussion of almost all the topics of the earlier edition of the book. Particular attention has been given to discordancy tests for univariate and multivariate samples and for data from structured models (linear models generally, and specific aspects of regression and designed experiment situations).
At a more general level, there has been a welcome growth of emphasis on methods of accommodation (robust inference in the face of outliers) and on informal (often graphical) descriptive procedures. These latter areas of development have contributed to the stimulus for a final type of modification - a reordering or reemphasis of some of the basic ideas and principles. In particular this has prompted a separation of general approach from specific results for the study of univariate samples with regard both to tests of discordancy and methods of accommodation, and a reversal of order in the treatment of these two aspects.

MSC:
62-02 Research exposition (monographs, survey articles) pertaining to statistics
62F35 Robustness and adaptive procedures (parametric inference)
62F03 Parametric hypothesis testing
62H15 Hypothesis testing in multivariate analysis