Testing the model of influence of presumed influence in a boundary condition: The impact of question order

We argue that evidence for and utility of the Influence of Presumed Influence (IPI) model lies in a robust causal chain in the form of self-exposure → other-exposure → perceived effects on others → behavior. A review of extant literature reveals competing theoretical explanations, as well as the possibility that a method factor (question order) could have driven the obtained models in structural equation modeling (SEM) analyses. The IPI model was tested in the boundary condition of different question orders using data from a web-based experimental study. The results from 2-group model SEM analyses show that the causal directions among presumed influence and media effects schema variables are indeed a function of question order.