For my first blog post on #CSEdAnalytics, I wanted to focus on two charts and one data point from the 2021 State of Computer Science Education Report that I feel are overlooked. A deeper analysis of this data shows how unevenly high school computer science is distributed throughout the US and provides valuable insights for those seeking to expand K12 computer science education within their state or nationwide.
On page 16 of the 2021 State of Computer Science Report, two charts hide in plain sight among other nationwide data. One shows the percentage of high schools with a foundational computer science course by geography, and another features this percentage by the free-or-reduced lunch (FRL) students in the school. I have recreated the data from these two charts below and added for comparison the national average of 51.3% of all high schools providing a foundational computer science (FCS) course.
Through the comparison with the national average, we can see two things:
- The only geographic category above the national average of 51% is suburban schools with 61%, and suburban schools surpass it by a significant amount. All other location/geographic types are below this national benchmark.
- Schools with student populations with FRL % < 50% have FCS courses above the national average, while schools with student populations with FRL > 50% are below the national average
To make it easier to express this information in fewer data points, I decided to create the following two metrics:
| Metric | Description | National Avg |
|---|---|---|
| Rural/Urban Relative Strength | (% of town and rural HS with FCS) / (% of urban and suburban schools with FCS) | 88.4% |
| Poor/Rich Relative Strength | (% of HS with FRL % > 50% with FCS) / (% of HS with FRL < 50%) | 76.5% |
One can interpret the two metrics above as follows:
- A student in a town/rural area is 11.6% less likely to have access to foundational computer science (FCS) courses than a student in a suburb/urban area
- A student going to a high school with FRL % > 50% is 23.5% less likely to have access to foundational computer science (FCS) courses than a student going to a high school with FRL % < 50%
Both of these metrics are significant. In contrast, the relative strength of Black/White student access to an FCS course is 92.7%, and the relative strength of Hispanic-Latinx/White student access to an FCS course is 96.0%. Future blog posts will discuss these metrics and others related to Black and Hispanic-Latinx populations.
Another data point mentioned on Page 16 of the 2021 State of Computer Science Education Report is:
“78% of US high school students attend a school that offers a foundational computer science course…”
While this is a generally favorable statistic, it does reveal that FCS courses are available primarily in schools with large student populations. Dividing the percentage of students attending a school with FCS (77.7%) by the percentage of schools nationwide that offer an FCS (51.3%) results in 1.515 or 151.5%. One can interpret this result as follows:
“The average population of a school with FCS (938) is 51.5% greater than the average school population (632).”
With a bit more math, we can calculate that the average population of the 48.7% of schools that did not provide an FCS course was 290, or 45.8% of the average school size (632). Or to combine it with the above result:
“The average population of a school with FCS (938) is 331% of the average population of a school without FCS (290).”
That current high schools with a foundational computer science course are significantly bigger, more affluent, and more suburban than those without is an inescapable conclusion of this deeper dive behind the 2021 State of CS Ed Report. This conclusion has important implications for those who want to expand K12 computer science education within their state or nationwide. My concern is that the easy part has been done. Expanding computer science education to the remaining 48.7% of smaller, poorer, more rural schools will be a more difficult “last mile” problem.
Getting computer science education to these smaller, poorer, more rural schools will challenge the US’s traditional public education infrastructure. Consider the actual difficulty of implementing some possible solutions below:
- Train many more teachers to teach one or two sections of CS in addition to other subjects
- Hire more teachers to teach CS at multiple schools/districts or part-time teachers
- Recruit more highly qualified teachers trained in CS Ed to teach at generally less prestigious smaller, lower-income, rural schools
- Learn to better leverage technology to reach schools that cannot afford to keep a dedicated computer science teacher on staff.
The solution to reach the remaining 48.7% will require considerable thought and effort by policymakers, education administrators, and teachers nationwide. We all wish them (and ourselves) great success in meeting this challenge.
*********
Please visit the CS Education Analytics page on this site for the underlying data and reports behind this blog post for more information. The attached reports will also show how your state compares to others in critical CS Education metrics. The next post in the #CSEdAnlytics series will investigate how female students fare in HS CS education.