Tableau School District Analysis

In my years as a public school teacher I spent lots of time analyzing student data to drive teaching decisions. I’d zoom out and look at the class as a whole to analyze how well they were mastering concepts. Then I’d zoom in and look at the data for each individual student so I could pinpoint the areas where they needed additional instruction or practice.

My principal frequently ran staff meetings that focused on school-wide data. But I thought it would be really interesting to dig into the data for an entire district. 

I came across this data set that was compiled from The Massachusetts Department of Elementary and Secondary Education (DESE). The data  includes all 1861 schools with over 950,000 students.

Of course I had questions…

❔ What are the graduation rates, and how widely do they vary by school?

❔ Does small class size yield better outcomes for college attendance?

❔ How much do test scores vary across the district?

You can view the entire Tableau dashboard I generated based on this data here.

Graduation Rates

I started my analysis by looking at graduation rates. The bar chart below provides a snapshot of the lowest performing schools. For the full picture, you can click here and scroll down further.

Graduation rate by school bar chart.

I was admittedly a little surprised by the variation. Yes, there are schools with 100% graduation rates. They are mostly charter schools, innovation schools, and academies. 

But I was really curious about the schools at the low end. The lowest performing school (Springfield Public Day High School) has a grad rate of only 8.7%. What could be going on with a school to yield such a low grad rate?

When I looked into the school stats, I discovered that 98.7% of the students are high need, 97.4% have disabilities, and the low income rate is 91%. And the school only houses 78 students — with a very high student/teacher ratio. Before I dug deeper, the graduation rate didn’t make sense. But while it doesn’t explicitly say so, the school is likely geared entirely towards very high-need students.

Class Size and College

Next I dug into class sizes. Ask any teacher and they’ll tell you that smaller class sizes yield better student performance. But I wanted to see if the data backs up that experience.

Class size, economic status, and college rates scatterplot

I looked for a correlation between smaller class sizes and students who attended college. And I was surprised. I created a scatterplot graph where each dot represents a school. I expected that smaller class sizes would correlate to higher graduation rates. But that’s not how it turned out. 

I’m not ready to accept that class size doesn’t matter. I started wondering what other factors might come into play so I added in data to reflect socio-economic status. The students in households that struggle financially did tend to be on the lower end of college attendance. That’s not a huge surprise, but it's good to confirm with the data. Generally, the schools with lower economic disadvantage rates had higher graduation rates.

If I were to dig deeper, I’d want to know how many of the schools with smaller class sizes are geared towards special needs students. As I mentioned above, Springfield Public Day High School has small classes and low graduation rates. But that can be explained when you consider that the school is likely geared entirely towards special populations.

4th Grade Math Scores

Since I was an elementary school teacher, I wanted to take a look at fourth grade math scores. I created a line graph that shows the 4th grade math test scores from low to high. I varied the color to mark the 50% point for clarity.

4th grade math scores by school

Again, I saw a big disparity between the lower-performing and higher-performing schools. I did a little digging into the bottom school. Hawlemont Regional is an elementary school that serves a rural area. The stats for the school show a 66.1% high needs rate, and a 53.6% low income rate.

Again, context matters when looking at school data.

Conclusions

Class Size and College Attendance

The data does not show a correlation between class size and college attendance. As a former teacher, my experience tells me that individual attention improves student performance. So what’s the disconnect? What other factors are impacting college attendance rates for the students who were in these smaller classes? I believe there’s more to this story.

Class Size and Economic Status

The students from lower-income households had lower college attendance rates. This isn't a surprise. Financial hardship certainly makes sending your child to college more difficult. Additional analysis on districts outside the state who have better low-income college matriculation might yield some data-driven interventions that would be worth studying.

Elementary Math Test Scores

Fourth grade passing rates vary widely across the state. Some schools rates are low enough to cause me to question the accuracy of the data. But once I did a little digging, I found some explanations for the low scores.

But — just like with graduation rates — digging into the data for the more successful schools to look for high-impact interventions that could benefit lower-performing campuses could be a game-changer.

But wait… There’s more

If you’d like to see another one of my data analysis projects, click here. 

I also love connecting with others in the data community. You can find me on LinkedIn. Feel free to reach out for a virtual coffee meet-up, a chat about data analysis and marketing strategy, or an opportunity you think I'd be a fit for!