Guided by the theoretical frameworks of interactive information searching and user engagement (UE), this study proposed sense discovery (SD) as a UE attribute and suggested a refined four-factor user engagement scale (UES) model for the measurement of users’ psychological involvement in web-based visual information searching. Using a mixed-methods approach based on a survey, this study confirmed the inter-item reliability of the original six-factor UES in three visual contexts—a general visual context, image searching on Google (ISG), and video searching on YouTube (VSY). Principal component analyses (PCA) partially confirmed the internal consistency of the original six UE subscales and suggested conceptual overlaps among four of six original subscales. Through thematic and sentiment analyses of the participants’ visual information needs, the study further explored their positive experience and categorized a total of eight items related to SD. Based on the findings, a refined four-factor UES model, which can be flexibly administered, is proposed to measure users’ psychological involvement in web-based visual information searching.