This study was conducted to optimize the designs of learning guides embedded in a computer-based simulation environment. The research was based on the Cognitive Theory of Multimedia Learning and Cognitive Load Theory. We investigated computer simulations under four conditions that combined representation and imagination learning strategies. This study recruited 244 fifth-grade Taiwanese students from various backgrounds (indigenous vs. nonindigenous, 98 vs. 146; boys vs. girls, 131 vs. 113) and examined the interactions and main effects of the four experimental conditions on the students' learning of geometric area concepts. The results revealed that the computer-based simulation platform can serve as a cognitive tool to help students to explore graphs and identify formulas. Moreover, the dynamic representations and trace changing imagination strategy embedded in the learning platform as guiding tools were highly beneficial for learners; this tool moderately reduced the learners' extraneous cognitive load and improved their learning performance at the application level. The static representation and concept imagination strategy embedded in the platform as a learning guide was also noted to increase the learners' internal cognitive load, which impaired their learning performance. Notably, boys and girls adopted different learning strategies, and the overall platform appeared to be more beneficial for nonindigenous students and students with middle to high levels of cultural stimulation than for indigenous students and students with low levels of cultural stimulation.