Effective teaching of conceptual modeling requires a carefully designed instructional approach due to the complexity of the learning tasks. To support learners, especially novice modelers, educators often use technology such as modeling and prototyping tools. This paper builds on this approach by using student-generated data to provide visual feedback on students’ learning activities, performance, and progress, delivered through a comprehensive unified interface called the Learning Analytics Dashboard (LAD). To facilitate interpretation of the LAD, students are provided with a frame of reference, often through peer comparison. While numerous studies have examined peer comparison in LADs, its effectiveness remains controversial. By providing both individual (default) and peer-referenced versions of the LAD in the context of conceptual modeling education, and closely tracking student engagement, performance, and LAD usage, this study seeks to uncover differences in the relationships between these factors for students from different academic backgrounds. The results reveal that these relationships are significant only for the peer-enhanced LAD, and they are stronger for students with less modeling experience, underscoring the value of peer comparison in LADs for novice modelers.