Mathematical communication skill was the missing piece of mathematics education failure puzzle for centuries despite its central role in learning activities. Built on top of other process skills, it was the tool to build more comprehensive conceptual understanding by learning math from different perspectives, sharpens other cognitive skills, and provides important feedback about students’ understanding. This study observed how the students’ mathematical communication skill was correlated with the application of generative learning model and how this correlation was influenced by various level of school’s academic achievement in mathematics. Using fractional 2x2-factorial design, careful non-probability sampling combined with simple random sampling to pick 171 students, and a validated and reliable scoring system to measure communication skill, this study proved that the students’ mathematical communication skill has a significant correlation with the application of generative learning model. The use of factorial design revealed that this correlation was more determined by the application of generative learning model, not by the level of schools’ or students’ mathematics achievement. Teachers should be more optimistic about the use of generative learning models in improving mathematical communication skill even in classes with lower level of mathematics achievement.

Keywords: mathematical communication skill, generative learning model, factorial experiment