The purpose of this paper is to provide an understanding of the issues related to formative modeling and the implications for its use, including some practical suggestions on methods for utilizing causal indicators within models.
In order to understand and explore the issues relating to formative models, this paper explores both formative and reflective indicators that together address the basic approach to the causal relationship within measurement models in structural equations modeling. Process recommendations for formative model/variable conceptualization and development are presented as well as a discussion of several evaluative issues including identification and proportionality. Some of the main arguments for and against its use are also discussed.
While the literature on formative models stresses its contribution to behavioral sciences, perhaps its greatest contribution is the flexibility of choice that it provides researchers in the development of models to more accurately estimate the relationship under consideration.
The paper contributes to the on-going debate on the relative advantages and disadvantages of the formative and reflective approaches to modeling and examination. Although the pendulum has swung in recent years away from an almost singular focus on reflective models, both approaches provide useful and complementary tools to theoretical management research modeling.
