Online peer-to-peer (P2P) lending platforms become a critical channel for financing. However, lenders suffer from economic lose due to dishonest behaviors of platforms and borrowers. By identifying risk correctly, lenders can make appropriate investment decisions to survive from risk and gain profits. Yet the existing empirical literature remains unclear about which factors influence lenders' risk perceptions and investment decision. Therefore, we draw on elaboration likelihood model to examine how lenders make their investment decisions away from problematic platforms and risky bids. Specifically, we identify argument frame as a central route of perceived risk, and platform reputation and source credibility as peripheral cues. We also examine roles of lender personal factors, i.e., cognitive capability and risk preference. We plan to conduct a laboratory experiment to collect data and test hypotheses. This study contributes to the cognition research in online financial products and provides insights for risk control and website design.