Abstract
Why bother about outliers? One can give several reasons. Outliers can mess up the statistical analysis. A few outlying observations can change the estimates of parameters of interest and their standard errors considerably. Many testing procedures we discussed in Chapters 2 and 3 and other procedures that are discussed below are based on the assumption of a constant multivariate normal distribution for the disturbances. The outcomes of these tests are hard to interpret if the normality assumption does not hold. One can confine oneself to so-called robust tests, see e.g. MacKinnon and White (1985), but their use is often only justified for null hypotheses that still do not capture some interesting outlier types. Moreover they can often only be applied sensibly in large samples, so that their appeal is largely theoretical.
This is a preview of subscription content, log in via an institution.
Buying options
Tax calculation will be finalised at checkout
Purchases are for personal use only
Learn about institutional subscriptionsPreview
Unable to display preview. Download preview PDF.
Notes
This chapter is a revised version of Ooms (1990).
Let \(\cdots,\xi_{-1},\xi_0,\xi_1,\cdots\) denote a strictly stationary sequence of random variables defined on a probability space (Ω, B, P). This sequence is said to be m-dependent if the random variables \((\xi_i,\cdots,\xi_k)\) and \((\xi_{k+n},\cdots,\xi_j)\) are independent whenever n>m, see Billingsley (1968, p. 167).
We use the original notation of Engle and Granger (1987).
Critical values 3.84, 5.99, 7.82, 9.49, and 12.59 for degrees of freedom 1, 2, 3, 4, and 6.
Backus et al. (1992) used these data without any outlier correction in their study on “International Real Business Cycles”.
See (5.6).
Author information
Authors and Affiliations
Rights and permissions
Copyright information
© 1994 Springer-Verlag Berlin Heidelberg
About this chapter
Cite this chapter
Ooms, M. (1994). Outliers. In: Empirical Vector Autoregressive Modeling. Lecture Notes in Economics and Mathematical Systems, vol 407. Springer, Berlin, Heidelberg. https://doi.org/10.1007/978-3-642-48792-7_5
Download citation
DOI: https://doi.org/10.1007/978-3-642-48792-7_5
Publisher Name: Springer, Berlin, Heidelberg
Print ISBN: 978-3-540-57707-2
Online ISBN: 978-3-642-48792-7
eBook Packages: Springer Book Archive