The ELM model developed by Petty and Cacioppo has been used in consumer behavior research for many years to explain how people form and change their opinions and attitudes towards a target. This study aims to find why there is no mention of bias correction in the high involvement of people in the ELM model. Furthermore, we explored a new moderator of argument diagnosticity in this study, and the strong diagnosticity argument was defined as being relevant to the true merit of the target and making consumers more likely to judge the product. The manipulation of high involvement situations and the comparison of four different argument qualities with strong and weak argument diagnosticity under two endorser conditions were employed. Bias corrections will occur when there is weak argument diagnosticity, indicating the argument diagnosticity is a crucial factor.