CodeBears is a summer computing camp at Morgan State University. I coordinated it across three summers, which means the version of it I know best is the operational one: recruiting, scheduling, matching mentors to groups, and the daily work of keeping a room of middle and high school students engaged with material that many of them arrived believing was not for them.
The research question underneath the logistics was about stereotypes. Not whether students could learn to program, which was never in doubt, but what they believed about who programs and who belongs in computing. Two research projects came out of the camp: one presented at SIGCSE 2024 examining stereotypes and their relationship to the STEM pipeline, and another presented at BICE 2024 exploring what the camp revealed about empowering underrepresented youth in computing.
What the camp taught me about teaching
Most of what I now believe about teaching came from this room rather than from a course. I often saw students disengage not because the material was beyond their ability, but because confusion could become evidence that they did not belong. That observation has shaped how I think about computing education: difficulty should be treated as part of learning, not as a signal that someone does not belong.
Everything I do in a classroom now is downstream of trying to break that inference.
That idea also follows directly into my current work with SensDS. CodeBears taught me to pay attention to the moments when students decide that something is not for them. With SensDS, I am interested in a related problem at a different level: what happens when we stop hiding the difficult parts of machine learning and instead make them visible and understandable? Rather than giving students a polished dataset and a model that appears to simply work or fail, SensDS lets them collect data, make decisions about that data, train a model, and encounter problems like distribution shift firsthand. In both cases, the goal is the same: make the hard parts of computing something students can engage with rather than something that quietly tells them they do not belong.
The associated poster and presentation received third place in the ACM Undergraduate Research Competition at the 2024 Tapia Celebration of Diversity in Computing.