A parsimonius approach to interaction effects in structural equation models: An application to consumer behavior

Structural equation models with non-linear constraints make it possible to estimate interaction effects while correcting for measurement error. Up to now, only direct effects have been specified, thus wasting much of the capability of the structural equation approach. In addition, the complexity of this approach has made its use much less appealing than it deserves in the marketing and management literatures. This article questions the actual usefulness of the constraints in the current specifications, and proposes reducing their number or even eliminating them completely, which leads to a more easily handled model that is also more robust to non-normality. The paper also presents and discusses an extension of Jaccard and Wan's, and Jöreskog and Yang's specification that can handle direct, indirect and interaction effects simultaneously. The approach is illustrated using empirical data for studying indirect and moderate effects of ?value orientations with respect to the environment? on both ?possibilities of influencing events? and on ?environmentally-friendly consumer behavior? such as buying and boycotting certain products.