Dr. Sheila Kohl
4.2.26
AI in K–12 education is a double-edged sword. On one hand, schools are trying to limit students’ use of AI for academic work. On the other hand, teachers are encouraged to use it to streamline time-consuming tasks such as creating rubrics, generating feedback, and analyzing data. My goal in the junior-level Science Methods course was to bridge the gap between the two and demonstrate to future educators how AI can be used in practice while allowing them to apply its results to an authentic task.
The methodology for this course:
1. Throughout the semester, I have shown students how I use AI to enhance my K–12 teaching practice—whether creating rubrics, designing worksheets, generating video review questions, or even planning an intramural sports competition for an active class. I rarely use AI-generated content as-is; instead, I treat it as a starting point, adapting and selecting the pieces that best fit my goals.
2. I am transparent with my class when I use AI in my SNC course. I show them the original prompt, walk through any revisions I made, and discuss the strengths and limitations of the final result—always emphasizing that I am ultimately responsible for the content I present. For example, in a choose-your-own-adventure activity, students took on the role of teachers and had to respond to a classroom situation involving a student who shot a spitball.
3. The major activity involved an assignment in which students compared AI-generated results to their own work. In pairs, students created and taught a lesson based on one of the Next Generation Science Standards Crosscutting Concepts. As part of the lesson, they designed an assessment for their classmates to complete to gauge their level of learning and understanding.
Later, students input their assessment components into different AI generators: a general-purpose tool (such as ChatGPT, Claude, or Gemini) and an education-focused tool (like MagicAI or School AI). Each student then compared all three outputs, evaluating the strengths and weaknesses of each in relation to their specific lesson.
To analyze the NGSS components, students completed a comparison matrix:
1. Comparison Matrix
| Comparison Questions | Self-Created Assessment | General AI Assessment (ChatGpt/Gemini/Claude) | Education Specific AI (SchoolAI, MagicSchool) |
| How accurately does the assessment align to the Disciplinary Core Idea (DCI)? | |||
| How well does the assessment measure the Science & Engineering Practice (SEP) (not just content knowledge)? | |||
| How clearly is the Crosscutting Concept (CCC) represented? | |||
| What level of thinking is required (recall, reasoning, application, explanation)? | |||
| How clear are expectations for student responses? | |||
| How developmentally appropriate is it for your selected grade level? | |||
| How well does the rubric/grading align with the purpose of your lesson? Does it clearly assess student understanding of the DCI? |
Next, they examined the data to respond to the following prompts:
A. Assessment Quality & NGSS Alignment
- Which assessment most effectively integrates all three dimensions- DCI, CCC, and SEP? How do you know this?
B. Comparing AI Tools
- How do the assessment tool results compare to each other?
- What strengths and weaknesses did you notice in each?
C. Teacher Judgment vs. AI Judgment
- Think about the types of students with diverse needs you may have in your classroom; is each assessment viable for all students, or will you need to make modifications?
- What professional decisions did you make that the AI tools did not?
- Did each of these create a usable assessment, in your opinion? Explain your thinking.
D. Conclusion- Future Classroom Implications
- Based on this experience, explain how you envision responsibly using AI to support (not replace) assessment design in your future classroom. (It can be any content area; not just related to Science).
Overall, I have found that while AI can spark ideas and make certain tasks more efficient, it cannot substitute for a teacher’s personal understanding of their students’ needs, the learning experiences they create, or the ways students best demonstrate their understanding.
Category: Resident Digital Fellow
