In order to explore the influence of signature form and rubric on the online peer assessment's engagement, this paper designed a two-factor mixed experiment of 2 (anonymous, real name) × 2 (with rubric, no-rubric) with 379 students from C College as the research object. The statistical analysis was carried out from the three dimensions of behavioral engagement, cognitive engagement and emotional engagement of learners participating in peer assessment. The study found learner's behavioral engagement in the real-name form is higher than the anonymous, and has nothing to do with whether or not to provide the rubric. However, in anonymous situation, the online peer assessment's cognitive engagement and emotional engagement with rubric are significantly higher than without the rubric. Therefore, it is recommended that teaching practitioners provide rubric when using anonymous peer assessment, but other relevant measures should be taken to improve learners' behavioral engagement.