Hierarchical linear modeling (HLM) has become an increasingly popular multilevel method of analyzing data among nested datasets, in particular, the effect of specialized academic programming within schools. The purpose of this methodological study is to demonstrate the use of HLM to determine the effectiveness of STEM programming in an Ohio middle school. This longitudinal study analyzes potential moderators of gender, socioeconomic status, student race, and attendance rate along with state test scores to quantify achievement. HLM determined integrated STEM education had a significant, positive effect on achievement in math and science combined (students scoring 31.8 points higher on average) and science achievement (students scoring 38.2 points higher on average) compared to traditional education students, respectively. There were little to no interaction effects determined between STEM participation and student factors. This demonstrates HLM as a powerful statistical tool for quantifying the impact of academic programming on student achievement.