2023 ASEE Annual Conference & Exposition

This study was conducted at a Research Experiences for Teachers (RET) Site in a university on the northern gulf coast. The RET site was funded by the NSF Division of Computer and Network Systems to offer a research-intensive program in artificial intelligence (AI) computing systems spanning application (cancer detection), algorithm, architecture and circuit, and device. Since the summer of 2021, Science-Technology-Engineering-Mathematics (STEM) middle- and high-school teachers participated in an annual six-week summer program. They participated in technology and instructional workshops, work sessions, and authentic artificial intelligence (AI) research activities with the university faculty, graduate, and undergraduate students. In the subsequent semesters, they developed and taught AI-integrated lessons in their classes.
Two designs of the summer program were implemented in 2021 and 2022. In the 2021 summer program, teachers participated in one of the four research projects, namely AI application in cancer detection, AI algorithm, architecture and circuit, and device. Teachers in the 2022 program learned AI algorithm and hardware foundational knowledge together before they were divided and spent more days on a specific AI research topic. Therefore, this study aimed to compare the impact of different RET summer program designs on participating teachers’ technological-content knowledge (TCK) and lesson plan development.
TCK Results
We used the Mann-Whitney U tests to examine the differences in participants’ technological-pedagogical-content knowledge between the 2021 (N=8) and 2022 (N=12) cohorts. Our results did not show a significant difference between the two cohorts in both pre-surveys and post-surveys.
We used the Wilcoxon signed-rank tests to examine the changes in teachers’ technological-pedagogical-content knowledge before and after the summer program in both cohorts. Results showed that the 2021 cohort did not have a significant improvement in their TCK (Z=-1.86, p=.06). However, the 2022 cohort’s TCK significantly improved at the end of the summer program (Z=-2.20, p<.05).
Lesson Plans Results
We analyzed 19 lesson plans developed by teachers at the end of the summer programs. A couple of differences were observed between and 2021 and 2022 cohorts. 1) The lesson plans developed by the 2022 cohorts demonstrated more AI integration in STEM classes. Those developed by the 2021 cohorts included concepts not on AI but relevant to the RET research projects, which could be a medical research topic or a foundational Computer Science or Engineering concept. 2) Teachers were adept at applying interactive learning technologies that explain AI working mechanisms to non-experts. The 2022 cohort learned to use the Teachable Machine and Thonny IDE. All of their lesson plans included the use of either or both technologies. The 2021 cohort learned python coding or hyperspectral imaging processing depending on their research groups. We also observed teachers bringing those technologies or knowledge to their classes.
Overall, the results of this study suggested a need to share the foundational AI concepts and learning technologies with all teachers if our purpose was to enable them to integrate AI into their classes.