Interpreting Interactions: Raw Means or Residual Means?

Ross, William T, and Elizabeth H Creyer

Journal of Consumer Research, 1993, vol. 20, issue 2, 330-38

Abstract: Interpreting interactions is important in analyzing data generated by the NOVA family of techniques. In this article, we examine how to specify the cell means used to describe an interaction. Although most researchers utilize the raw means that are the output of most analysis packages, R. Rosenthal and R. L. Rosnow suggest an alternative procedure, removing the treatment effects (and possibly the grand means) from the group means and then examining the interaction in isolation. We examine why and under what circumstances it is appropriate to use residual means to examine interactions. How the use of raw means versus residual means affects interpreting interactions is examined in the context of a demonstration experiment. After that, a procedure for interpreting interactions is presented. Copyright 1993 by the University of Chicago.

Date: 1993
References: Add references at CitEc
Citations: View citations in EconPapers (3)

Downloads: (external link)
http://dx.doi.org/10.1086/209353 (application/pdf)

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:oup:jconrs:v:20:y:1993:i:2:p:330-38

Access Statistics for this article

Journal of Consumer Research is currently edited by Bernd Schmitt, June Cotte, Markus Giesler, Andrew Stephen and Stacy Wood

More articles in Journal of Consumer Research from Journal of Consumer Research Inc.
Bibliographic data for series maintained by ().