HS Computer Science in Tennessee: Analysis of AP CSP Performance Over a 4-Year Span

By: Mehreen

A few months ago, I wrote an op-ed for the Tennessean highlighting the importance of investing in diverse tech talent in Nashville. Not long after, Oracle announced their new office in Nashville, along with the arrival of 8,500 jobs by 2031 — applying more pressure for schools, companies, and bootcamp programs to meet our city’s growing hiring demands and to increase efforts to build a talent pipeline in Tennessee.

While other programs targeting high school learners exist across the nation — such as Code NationAmerica On TechUpperline CodeCode2CollegeOperation Spark, and CodeCrew to name a few — very little resources have been invested to creating similar opportunities for the high school market in Tennessee, and more specifically middle TN. If Nashville is to become the tech epicenter of the South, we must invest more resources, time, and funding in supporting K-12 initiatives focused on diversifying the tech talent pipeline and equipping our young people with industry-aligned skills.

I’ve recently been working on launching a program called Culturally Tech with a few friends — Stephen Castaneda and Namita Manohar. We are New Normal Fellows with 4.0 Schools. You can read more about our project as well as other education innovators here — website coming soon🤓. In creating our program’s mission and vision, we wanted to get a better understanding of the current state of CS access, exposure, and persistence for high school students in Tennessee.

One of the most important metrics we use at the high school level is the AP Computer Science Principles’ passage rate. The AP CSP exam is a relatively new exam — officially launched in the 2016–2017 SY — and seeks to introduce students to core computer science concepts before taking a more advanced course like AP CSA. Given that there is no prerequisite for the course and that it has increased in popularity among AP students, it’s a good metric to compare performance across regions and demographics. All archived data is available to the public and can be accessed here. I pulled raw data 📊 for the AP CSP exam from the College Board for Tennessee from the past four years and used it to calculate and compare a number of other passage rates.

Data is calculated to show the number of Tennessee AP CSP test-takers passing by demographic and year.
Graph representation from 2017–2020 by demographic of passing test-takers. In general, more students are enrolling in AP CSP and more students are passing the exam.
Data is calculated to show the percentage of passing test-takers by demographic out of total test-takers in each demographic. For example, roughly 80% of Asian-identifying test-takers passed the exam in 2020 compared to roughly 44% of Black-identifying students who sat for the exam.
Graph representation of performance within student demographic.

In 2020, there was a sizable drop in passage rates across demographics, with the exception of Black students and multiracial students. This was particularly striking given that Black and Latinx communities were disproportionately affected by Covid-19 in a number of ways. What contributed to that 13% jump for Black-identifying students? 📈

My first thought was that the strong performance by RePublic High School (Nashville, TN)— an AP CSP-For-All school where 90% of the student body identify as students of color — was the contributing factor. RHS doubled its passage rate of students from 37 students in 2019 to 75 students in 2020. While RHS students represent about 8% of the overall passing TN test-taking population, roughly 35% and 30% of Black and Latinx passing TN test-takers respectively attend RHS. This is significant and worthy of discussion, yet still doesn’t explain the increase as RHS students have — with some fluctuation — represented approximately a third of the state’s passing percentage within the Black demographic over the past 4 years.

During the pandemic, the College Board modified the AP CSP exam, eliminating the multiple-choice exam portion of a student’s score. Instead, students just submitted the Create and Explore Task projects. This makes me wonder about the impact of project-based learning and other ways in which we can assess student understanding. While standardized tests are a common metric to compare student performance, it can also be a gatekeeper for highly-qualified students to being accepted into the college or university of their choice. The AP CSP exam was further modified in 2021, dropping the Explore Task and increasing the overall weight of the multiple-choice exam portion in calculating student scores. While individual student scores and school reports are not scheduled to be released until late July, I am curious to see how these modifications impact student performance across demographics, and potentially reverse the growth we saw from 2020.

Percentage of each demographic passing compared to the overall passing numbers. For example, White students represented 53.40% of the passing TN test-taking population in 2020.
Graph shows the percentage of each demographic in comparison to all passing demographics.

One of the strongest AP CSP curricula available is developed by Code.org. The non-profit organization has led in K-12 efforts to increase more girls and underrepresented students pursuing computer science through designing teacher workshops, creating professional development groups, launching the Hour of Code initiative globally, and leading advocacy campaigns. In 2020, Code.org published the following stats specific to TN based off of surveys from their student community:

“Students who learn computer science in high school are six times more likelyto major in it, and women are ten times more likely.

“Black students are more interested in CS and more confident in their abilities than white students but are less likely to attend a school that offers it.”

In designing Culturally Tech programming, our team aims to leverage these two findings by doing the following:

  1. 🔗 Connect students to opportunities and selective programs that exist outside of traditional school time and reduce the barrier to participation by providing student stipends.
  2. 🌐 Ensure that more underrepresented students have access to a network of peers and mentors interested in CS as well as industry-aligned technical opportunities should they attend a school without a CS program.
  3. 👩🏽‍💻👨🏿‍💻 Build a community for HS students in middle TN — with plans to expand — through both in-person and virtual programming.

If you wish to get involved, let’s connect. We are looking for teaching assistants, mentors from a variety of tech backgrounds, guest speakers, and anyone passionate about supporting the future generation of technologists. If you know of any 8th-12th grade students interested in participating, direct them to this form. Stay tuned for the 🚀 launch of our website and details about our July 31st pilot.

Leave a Reply

%d bloggers like this: