Christina Gardner-McCune

This article provides an in‑depth look at how K–12 students should be introduced to Machine Learning and the knowledge and skills they should develop. After an overview of the AI4K12 Initiative and its “Five Big Ideas in AI,” the paper explains the guidelines’ structure and grade‑band progressions, presents a theoretical framework for developmental appropriateness, and unpacks the Learning (Big Idea 3) progression across grades with horizontal and vertical examples. It concludes with guidance on creating learning experiences that connect across the Five Big Ideas and free online tools that support them. citeturn8search1

In introductory CS courses, students typically program in text‑based languages, requiring simultaneous mastery of syntax and semantics. Dual‑modality programming environments link blocks and text representations, potentially easing this burden. In a study across two semesters of a CS1 course (N=673), students taught with dual‑modality representations in lecture and provided dual‑modality tools performed better on typical course exams than students in a text‑only setting, suggesting dual‑modality instruction can improve early CS learning. citeturn19search0turn19search1

Industry internships offer CS students authentic disciplinary experience and pathways to employment, yet little is known about participation patterns and preparation processes. Surveying 536 CS undergraduates at three U.S. universities, the study found about 40% completed at least one internship. Participation did not differ by GPA or gender, but higher socioeconomic status correlated with interning. Interns explicitly prepared (e.g., interview practice), while non‑interns relied on coursework or were less engaged in applications. Results suggest factors outside coursework strongly influence internship access and outcomes. citeturn15search0

This study presents validity evidence for using the S‑STEM instrument to evaluate changes in attitudes within a middle‑school robotics learning environment. Data from students participating in a district‑wide robotics integration effort provided internal‑structure and criterion‑related validity evidence. Measurement invariance analyses indicated S‑STEM items functioned equivalently across groups (e.g., grade level). The results support S‑STEM as a useful tool for assessing attitude changes in STEM programs and offer implementation guidance. citeturn23search0turn23search1

The ubiquity of AI in society makes it timely to consider what educated 21st‑century digital citizens should know about AI. In 2018, AAAI and CSTA formed a joint working group (AI4K12) to develop national guidelines for teaching AI, machine learning, and robotics in grades K‑12 and to curate an online resource directory for teachers. This Blue Sky paper invites the AI research community to reflect on “big ideas” in AI every child should know and to help communicate about AI advances and their societal impacts. It calls on more AI researchers to create resources that enable teachers and students to understand AI. citeturn6search0