2026 ASEE Annual Conference & Exposition

The increasing adoption of AI in education has encouraged educators’ and learners’ interest in this transformative technology. However, both learners and educators experience AI mostly as a software application, typically in the form of a chatbot, trivializing AI. There is also a growing need to explore more accessible and context-aware applications of AI, one of which is Edge AI. As computer engineering becomes increasingly important in our society, it is essential to understand how young learners engage with microelectronics and Edge AI. This study, funded by National Science Foundation, seeks to contribute to our collective knowledge regarding this gap and explore the evolution of informal learners’ situational interest in microelectronics and edge AI.

This study was conducted during a two-week summer program at an innovation museum in the Southeastern US and was guided by the Four-Phase Model of Interest Development (Hidi & Renninger, 2006, 2022). It also incorporated the engagement framework by Fredricks et al. (2004), which conceptualizes engagement as cognitive, emotional, and behavioral. The program involved 11 middle-school aged summer camp participants (aged 11-13) using a five-module curriculum: (1) embedded systems, (2) data acquisition, (3) multimodal sensing, (4) AI and Edge AI, and (5) Edge AI for social good. The curriculum utilized an AHA! board which uses a variety of sensors for multimodal data collection, a powerful ESP32 microcontroller, an OLED screen and RGB lights as well as a breadboard powered with Arduino IDE and Edge Impulse.

A mixed-method design combines quantitative and qualitative approaches. Quantitatively, participants completed the pre- and the post-Situational Interest Survey (SIS, Linnenbrink-Garcia et al., 2010). Qualitatively, data sources included daily reflections, interviews, and structured observations analyzed using Braun and Clarke’s (2006) six-phase thematic analysis. Inter-rater reliability (IRR) was assessed using the percentage agreements. The two-observer configuration yielded an average agreement of 97%; SD=0.09 while the three-observer configuration achieved 99%; SD=0.09, indicating strong consistency. Interview responses and daily surveys enriched the analysis by capturing learners’ engagement.

Only six participants (4 girls, 2 boys) completed all quantitative instruments so results are presented descriptively to support qualitative findings. Pre/post SIS scores showed very modest gains in overall situational interest (Mpre=13.1, Mpost=13.3) and perceived value (Mpre=12.8, Mpost=13.2), while maintained feeling decreased slightly (Mpre=13.6, Mpost=13.2). Qualitative data revealed that hands-on, collaborative, and creative tasks (e.g. PC assembly, AI drawing, sensor prototyping) consistently triggered situational interest. Cognitive engagement indicators such as creativity and understanding peaked early during project design. Emotional engagement data showed boredom spiking during facilitator talk and pride peaking during project work and final presentations. Behavioral engagement reflected high participation during hands-on activities.

The full paper will detail curriculum artifacts, observation protocols, SIS survey, and thematic analysis. It will also provide implications for implementation of AI and microelectronics in informal learning contexts and for pre-college engineering pathways.