You can look at experience as the tip of the iceberg; which is why it is the top of the CAPE Triangle. Experience is what most people see when they look at data.
To investigate experience, we will use data to answer the following questions:
1) What are our AP CS passing rates, over time? This data is parsed to show district wide results and results specific to various race and gender subgroups.
2) When asked, how do students describe their experience in CS classes?
Below, you will find two sets of data sets organized by data source.
Key Conclusions:
Referring to the Passing Rates data, you can see West Hartford, State, and National data that confirms that passing rates for AP Exams differ between student subpopulations based on demographics that should not be correlated with academic achievement.
For example, looking at AP CS A, in 2020, our Asian/Pacific Islander population pass rate is 74%, while our African American/Black population pass rate is 60%.
Looking at AP CS Principles in 2020, our Asian/Pacific Islander population pass rate is 100%, while our African American/Black population pass rate is 57%.
These variations in pass rate on the same assessments raise concerns.
Referring to the Passing Rates data, you can see comparisons of subgroups, with local, state, and national data alongside each other. Note that high passing rates can be caused by too many students opting out of taking a course. High passing rates (such as the high passing rate of females in both AP CS A and Principles and our latinx/hispanic population in AP CS Principles) suggests the underrepresentation of both in computer science courses.
When students opt to take CS at the high school level, the reason they do so varies.
There is not one solution to recruiting more students and/or achieving diversity in computer science. Indeed, it seems that a hybrid approach that combines a presentation in the fall and teacher and counselor conversations during course recommendations is the best approach.
It is important that all students see the connection that CS has to career, student personal interests, and it is important that students hear their peers sharing these ideas.
The reasons students take CS at the high school level varies by students' cultural/racial identification.
Approximately 29% of Asian/Pacific Islander and 35% of White students report that Teacher/Counselor conversation was one of the reasons that they signed up for computer science. If we compare this to 42% of Black students and 43% of Female students, two minority populations in CS, we can see the necessity of these conversations for underrepresented minorities in computer science.
Least likely to name "someone outside of school" as the reason they signed up for computer science coursework were Latinx/Hispanic and African American / Black students.
Least likely to name "career" as the reason they signed up for CS coursework were Latinx/Hispanic and Females, especially Latinx/Hispanic and Black Females.
Student perceptions of their experience in AP and non-AP courses are overwhelmingly positive. The majority of students in AP and non-AP courses feel like they belong, feel liked by their peers, feel accepted, do not feel like an outsider and do feel included. However, there are discrepancies in experience when looking at specific demographics.
In AP CS classes, African American/Black and Latinx/Hispanic students are 20-25% less likely than White and Asian / Pacific Islander students to feel accepted and included.
When the full survey results were collected, 80-100% of the 225 students surveyed reported feeling like they belong in CS. However, only 60% of our Latinx/Hispanic Females and Black Females believe that they are liked by their peers.
Drilling down into the data further, you will see the differences across schools and in non-AP courses.
Gender inequities in student experience exist across all CS courses. This is demonstrated via the perception vs. reality charts shown for AP courses and across all HS courses. Students were asked how much they thought they understood CS when compared to their peers. Students were then asked their grades. Consistently, our female students reported that they thought they were understanding less than their peers. Males reported that they thought they were understanding more. When compared to letter grades, most of the time, female students had the higher grade.
This gap between perception and reality exists in both AP CS A and AP CS Principles.
Looking at AP CS A, males were 7% more likely to report that they understood CS better than their peers, but the grades were within 1% of each other.
For AP CS Principles, males were 25% more likely to report that they believe they understand CS better than their peers, but their reported letter grades were 7% lower than females.
Looking at all 225 students surveyed across all CS courses, males were 22% more likely to report they understand CS better than peers. Females reported about 9% higher in terms of grade performance.
This can lead females to underestimate what they can take in terms of courses and that lack of confidence may cause them to opt out of computer science or not take the most challenging courses in computer science.
We are seeing evidence of this in our AP CS Principles and AP CS A passing rate data where females are outperforming all other subgroups.
This suggests that only the most academically confident females are opting in to CS Courses.
Attendance issues also indicate student experience. For the attendance charts here, students self-reported the amount of times they have missed class, been more than 15 minutes tardy, or purposefully skipped class. African American/Black students were two times more likely to report missing 1-2 classes, skipping 1-2 classes, and be 15 min tardy 1-2 times than their peers. 9% of females reported missing 1-2 CS classes in the past two weeks.
[Copyright 2021 West Hartford Public Schools, Email jackie_corricelli@whps.org with comments or questions.]