Source and message factors in persuasion: A reply to stiff's critique of the elaboration likelihood model

Abstract In this article we respond to James Stiff's (1986) recent critique of the Elaboration Likelihood Model (ELM) of persuasion (Petty & Cacioppo, 1981, 1986b). In particular, we make the following corrections to Stiff's misrepresentation of the model: (1) Many variables other than “involvement”; can affect the elaboration likelihood and thus the route to persuasion, (2) variables can serve in multiple roles under specifiable conditions, and (3) the ELM does not preclude multi‐channel information processing. After correcting these misperceptions of the ELM, we critique Stiffs meta‐analyses comparing the ELM predictions with those he derives from Kahneman's (1973) elastic capacity model. His analysis of message factors is critiqued on the grounds that some of the message factors included in the analysis are capable of affecting attitudes via either the central or the peripheral route. His analysis of source factors is critiqued on the grounds of insufficient sample size, lack of statistical significance, and possible miscategorization of studies. In short, we argue both that Stiff's presentation of the ELM and the conclusions he draws based on the data he presents are misleading.