Students' perceptions of AI are often shaped by commercialized and fictional depictions of AI, leading to misuse, underuse, and over-reliance on AI tools. Such preconceptions place additional burdens on K-12 AI curricula to not only educate students on foundational AI knowledge components, but also uncover and reshape students' understanding of AI. To understand ways to uncover and address students' AI misconceptions in formal learning settings, we conducted a study that taught foundational AI concepts using a formal AI curriculum, games, interactive activities, and scenario-based challenges to high school girls. We analyzed student dialog and survey responses data to observe shifts in their perceptions and understanding of AI. Our findings indicate that unstructured AI activities effectively scaffold student reflection and metacognition on AI topics, contributing to significant shifts in perception. This work provides insights for designing engaging and effective AI curricula that address students' misunderstandings of AI.

The increased demand for Computer Science (CS) skills across most professions has necessitated providing CS training for post-secondary learners regardless of their fields of study. This is especially true in developing countries with high unemployment rates, where workers need to update their skills continuously to remain employable. Understanding the motivations of students who enroll in CS courses with limited prior training is critical for designing learning environments that help them achieve their desired goals. This paper studied the motivations of 2376 African students and working professionals enrolled in a smartphone-based programming MOOC. We found that the non-CS major participants wanted to learn programming to develop business applications, keep up with the demand for computing skills in their careers, and produce tools that positively impact their community. We also found that factors such as employment status and participants' motivations influenced students' persistence through the course. Our research contributes valuable insights for accessible and effective programming MOOC design for African professionals, especially for those without CS backgrounds.

Computer science education in Africa is gaining momentum as computer science skills give employees a competitive advantage in an increasingly globalized labor market. Researchers and educators have addressed this demand by making CS instructional materials more accessible, creating professional development (PD) resources, and designing online courses targeted at undergraduate CS education settings. However, there is limited research on what types of support African teachers need to improve pedagogy in K-12 settings, which is critical for fostering foundational interest in CS. To address this, we interviewed fourteen CS teachers in nine K-12 schools in the Greater Accra Region of Ghana to uncover opportunities to design interventions that address challenges they experience with teaching CS in schools and support their professional development. We found that most CS teachers lack formal CS training and therefore taught students self-taught topics from their personal education or work experiences, regardless of the requirements of the government curriculum. We also found that CS teachers are by default assigned computer-based school administration responsibilities, such as grade administration and managing school websites, further limiting time for professional development. We propose solutions for creating professional development tools considering these teachers' CS experience, limited computing resources, professional development time, and curriculum requirements.

The sine illusion is an underestimation of the difference between two lines when both lines have increasing slopes. We evaluate three visual manipulations on mitigating sine illusions: dotted lines, aligned gridlines, and offset gridlines via a user study. We asked participants to compare the deltas between two lines at two time points and found aligned gridlines to be the most effective in mitigating sine illusions. Using data from the user study, we produced a model that predicts the impact of the sine illusion in line charts by accounting for the ratio of the vertical distance between the two points of comparison. When the ratio is less than 50%, participants begin to be influenced by the sine illusion. This effect can be significantly exacerbated when the difference between the two deltas falls under 30%. We compared two explanations for the sine illusion based on our data: either participants were mistakenly using the perpendicular distance between the two lines to make their comparison (the perpendicular explanation), or they incorrectly relied on the length of the line segment perpendicular to the angle bisector of the bottom and top lines (the equal triangle explanation). We found the equal triangle explanation to be the more predictive model explaining participant behaviors.