The use of simulations to better convey abstract scientific concepts or difficult-to-observe scientific occurrences through the manipulation of variables or virtual objects has the potential to improve student science learning. However, misconceptions or extraneous load may emerge if simulation support is insufficient. In this study, we used a self-explanation strategy via question-based prompts in an online scientific simulation learning environment to improve the comprehension of junior high school students in physics concepts, engagement and motivation toward science learning, and task load. Thirty-seven seventh grade students from two junior high school classrooms in northeastern China were allocated into experimental and control groups using a quasi-experimental approach. Students in the experimental group utilized a simulation-based learning environment that incorporated self-explanation, while students in the control group used the same environment but without self-explanation integration. Additionally, sequential analytic approaches were used to examine the behavioral patterns of both groups of students throughout simulation operations and the learning process. The results showed that the students in the experimental group demonstrated significantly better learning achievement than those in the control group; furthermore, we discovered that the students in the experimental group displayed more inquiry behaviors regarding prediction, simulation operation, and explanation than did those in the control group.