Where Have all the Moderators Gone: The Perils of Type II Error.
Abstract
Traditional moderated regression analysis requires a significant and sizable increment in R2 due to the interaction term in order to conclude that there is a moderator effect. A Monte Carlo analysis demonstrates that the traditional procedure will not discover moderators even when they are the only existing effect. A suggestion is made to reverse the usual hierarchical order of the moderated regression analysis. The choice between a linear and an interactive model should be made on the basis of theoretical justification. (Author)
Document Details
- Document Type
- Technical Report
- Publication Date
- May 01, 1977
- Accession Number
- ADA040598
Entities
People
- Gary M. Mullet
- Milton R. Blood
Organizations
- Georgia Tech