The explosive growth of misinformation on social media has a great negative impact on public health, democracy, and social cohesion. To detect misinformation on social media, we examine which types of textual features are more effective in supporting misinformation detection on social media platforms. Drawing on the Elaboration Likelihood Model (ELM) of persuasion theory, we incorporated the features of misinformation into two routes of the ELM: the central route and the peripheral route. We will use econometric modelling techniques to evaluate the effects of different features (support central and peripheral cues) on social-media users’ perception of news authenticity. Our research is expected to contribute to the current body of knowledge related to misinformation detection in two ways: a) facilitating explainable machine learning models for detection, and b) exploring the effect of the persuasion process on users’ perception in this context. Our research is expected to make significant contributions to practice.