Math Test Results – New York City, Grades 3 to 8
Idea & Approach
The provided dataset contains math test results for children in New York City, from grade 3 through grade 8 — roughly ages 7 to 14. The results are grouped in various ways, for example by whether students have ever taken English language lessons or have a learning disability. For each of these categories, both a median score and a proficiency level are reported.
There are four levels in total:
- Level 1: insufficient basic skills
- Level 4: above-proficient skills
My idea was to explore the differences between groups. Initially, I visualized the data in a Dash/Plotly app, and later rebuilt it in Plotly Studio — partly due to time constraints and a bit of a “what now?” moment.
Data Enrichment
The provided dataset covered the years 2013–2023. Additional data was available for 2006–2012, which I merged successfully for most categories.
After combining the datasets, a significant drop in performance became visible around 2010 and 2013 — more red, less blue.
The main reason appears to be a change in testing approach: greater emphasis on reasoning and problem-solving rather than basic skills. To exaggerate slightly — before, students were simply asked to solve 4 × 9, whereas later they had to read a short story and figure out on their own that 4 × 9 was the calculation needed.
Result
From my own app, I discovered — to my surprise — that children who once received English language lessons performed better than those who are currently in such programs or never received them. This pattern wasn’t just incidental; it held true regardless of how the data was analyzed.
In the image below (right side, ELL stands for English Language Learners), the more blue and green, the better the results. Combined with the testing changes introduced in 2010 and 2013, it raises an interesting question: at what point does testing math skills start to overlap with testing language proficiency?
This is also visible when looking at the results in Plotly Studio. The Plotly Studio app — which took about an hour to build and fine-tune — offers four different perspectives for exploring and filtering the results. The images below show the charts created in Plotly Studio.
Plotly Studio charts
Figure Friday is an initiative by the Dash/Plotly community, where each Friday a new dataset is shared and participants create a visualization or small app to extract insights from it. The following Friday at 18:00, there’s a Zoom session where some participants explain the thought process behind their work. In the community thread, people also share their code — and when possible, a live demo — to learn from each other.