M.S. Applied Data Science - Capstone Chronicles 2025

9

Scatterplot: Chronic Absenteeism vs Graduation Rate

There was insightful information seen from instructional indicators. Schools that had higher percentages of experienced teachers tended to have higher graduation rates. However, schools that had first-year teachers or inexperienced teachers had lower outcomes. There was also a slightly modest positive relationship between student-teacher ratio and graduation rates. In this dataset, it was revealed that schools with large class sizes tended to have slightly higher graduation outcomes. This pattern may reflect a broader contextual factor, such as school size, district resources, or community demographics. The feature-target correlation supported the visual patterns displayed. Dropout rate and still enrolled rate had the strongest negative correlation with graduation ( r =- 0.91, -0.67

respectively), this is because they directly represent students who did not graduate. To continue, the strongest negative predictors were unexcused absences ( r = -0.47), chronic absenteeism ( r = -0.43), and FRPM eligibility ( r = -0.44). As for the strong positively correlated variables, they were not as high in number but are still notable: A –G completion rate ( r = 0.37), student–teacher ratio ( r = 0.25), and the percentage of experienced teachers ( r = 0.18) were the ones with the largest positive correlation. Safety and climate indicators resulted in small positive correlations which suggest that these variables have an effect on graduation indirectly or through a broad school-level condition as opposed to a direct relationship.

198