Exploring Interaction Effects in the Anova

Morton B. Brown

Journal of the Royal Statistical Society Series C, 1975, vol. 24, issue 3, 288-298

Abstract: In a multiway anova the interaction term may result from the presence of one or more outliers. Another explanation may be that the region of applicability of a more parsimonious model is limited and does not include the whole region spanned by the observations. In both cases the identification of the cells whose means have large residuals when their expected values are computed by the simpler model can be an aid in drawing inferences about the data. A stepwise routine is suggested for the detection of cells which are major contributors to an interaction sum of squares. At each step the expected values of the cells previously selected are calculated by fitting to the remaining data the model which does not contain the interaction term being studied. The differences between these cell means and their expected values according to the model are then used in drawing inferences.

Date: 1975
References: Add references at CitEc
Citations:

Downloads: (external link)
https://doi.org/10.2307/2347089

Related works:
This item may be available elsewhere in EconPapers: Search for items with the same title.

Export reference: BibTeX RIS (EndNote, ProCite, RefMan) HTML/Text

Persistent link: https://EconPapers.repec.org/RePEc:bla:jorssc:v:24:y:1975:i:3:p:288-298

Ordering information: This journal article can be ordered from
http://ordering.onli ... 1111/(ISSN)1467-9876

Access Statistics for this article

Journal of the Royal Statistical Society Series C is currently edited by R. Chandler and P. W. F. Smith

More articles in Journal of the Royal Statistical Society Series C from Royal Statistical Society Contact information at EDIRC.
Bibliographic data for series maintained by Wiley Content Delivery ().