Designing for Sensemaking: AI, Rigorous Assessment, and Making Student Thinking Visible
Reflections from the 2026 Colorado Science Conference
November 6, 2026 | MosesSTEMinc L3C
Today, I had the opportunity to attend an excellent session at the Colorado Science Conference titled Designing for Sensemaking: Using AI to Strengthen Instruction and Assessment in K–12 Science, Technology & Engineering, presented by Samantha Agoos, Science Content Specialist with the Colorado Department of Education. I appreciated that this session focused on designing rigorous science assessments that allow students to engage with artificial intelligence while still making their thinking visible. The session was engaging, interactive, and provided relevant examples of how AI can support science instruction without replacing the thinking students need to do.
Looking at Student Work: What Do We See Students Thinking?
One of my favorite parts of the session was the opportunity to analyze student work samples from OpenSciEd. The presenter asked an important question:
“Where do we see student thinking?”
I really liked this part of the session because we were not simply looking at whether students had the correct answers. We examined student models and discussed what their work revealed about their understanding of scientific concepts. It was an opportunity to think about how students represent their ideas, use evidence, and communicate their understanding. As science teachers, we spend a lot of time planning instruction and developing assessments. However, examining student work is just as important. What are students telling us through their models, explanations, and responses? What do they understand? What questions do they still have? These are the kinds of questions that help us make instructional decisions based on what students are actually thinking.
Sensemaking Is a Process
Another part of the session I loved was the discussion about the sensemaking process. The presenter used an excellent visual from Aneesha Badrinarayan's book, Surfacing Brilliance Through Meaningful Science Assessment: Shifting Practice to Reflect Today's Students and Today's Standards (2025), to illustrate how students develop scientific understanding. The visual showed how students bring prior knowledge and experiences into learning, encounter uncertainty, engage in scientific reasoning, and gradually develop ideas they can apply to new situations. I especially appreciated the discussion about uncertainty. Students do not always have an immediate explanation for what they observe. Sometimes they need opportunities to work through their ideas, examine evidence, ask additional questions, and revise their understanding.
The four elements of meaningful science assessment discussed during the session were:
Sensemaking
Relevance
Student Assets
Opportunity
I appreciated how these elements were connected to assessment design and the importance of making student thinking visible (Badrinarayan, 2025). The visual also reminded me that sensemaking is not always a straight path. Students need opportunities to investigate, make connections, reconsider their ideas, and develop explanations over time.
Assessment Happens Throughout Instruction
Another important point discussed during the session was that assessment should happen throughout instruction, not only at the end of a lesson. The presenter shared a sequence that connected learning tasks to instructional decisions:
Task → Student Thinking → Evidence → Teacher Response
I really liked this because it places student thinking at the center of assessment. When students complete a science task, their responses provide evidence of what they understand. Teachers can use that evidence to determine what instructional support or additional learning opportunities students may need. This becomes especially important when we incorporate AI into science assessments. We still need to ask:
Where is student thinking visible?
What evidence are students providing?
Are students making scientific connections and developing explanations, or is the technology doing the thinking for them?
These are important questions to consider as AI becomes part of classroom instruction and assessment.
Ghost Cats: A Rigorous and Engaging Science Investigation
One of the assessment examples I particularly enjoyed was Ghost Cats: Tracking Bobcats in the Sagebrush, a high school science task aligned to HS-LS2-4. The investigation focused on Colorado's bobcat population and the relationship between human harvesting, predator-prey interactions, energy flow, and matter cycling in a sagebrush ecosystem. Students were asked to analyze population data, compare study sites, complete mathematical calculations, develop ecosystem models, and use evidence to support scientific explanations. I thought this activity was excellent! It was engaging, rigorous, and culturally relevant to the Colorado setting because it connected science instruction to local wildlife and environmental issues. I appreciated that students were expected to do more than recall scientific information. They had to examine data, identify patterns, develop models, and explain how changes in one part of an ecosystem could affect other organisms. The activity also connected scientific investigations to careers in wildlife science.
This is definitely an activity I would consider using with my own Biology students.
I could see opportunities to connect this investigation to ecosystems, population dynamics, matter cycling, and energy transfer.
Using AI Without Replacing Student Thinking
Another aspect of the session I appreciated was the discussion about allowing students to use a customized Gemini Gem during science learning. I liked seeing how AI could be intentionally incorporated into rigorous science assessments while students remained responsible for analyzing evidence, developing models, and explaining their scientific reasoning. Teachers can also create customized Gemini Gems and assign them through Google Classroom, depending on their school's access and settings. I wonder how many educators know about this feature. Rather than focusing only on whether students should be allowed to use AI, I think we should also examine how we design the tasks students complete.
Are we designing assessments that require students to reason, apply scientific concepts, evaluate evidence, and communicate their understanding?
Can students use AI as a resource while still demonstrating what they know?
AI can support students during learning, but it should not replace their scientific thinking.
For me, this was one of the most important connections between AI and rigorous assessment design.
Exploring AI Tools for Science Teachers
I was also impressed by the AI tools introduced during the session. The presentation included resources such as Science Standards Unpacked, the NGSS Phenomena Generator, and the NCDPI Judgment and Agency Architect. These tools offer opportunities to examine science standards, explore phenomenon-based tasks, and consider how AI can support instructional and assessment design. Although these tools were introduced during a session presented by a Colorado Department of Education specialist, they were not all identified as tools developed by the Colorado Department of Education. What I appreciated was seeing relevant ways teachers could intentionally integrate AI into instructional planning and assessment design while maintaining a focus on student learning.
Learning Across Grade Bands
Another feature of the session I appreciated was the inclusion of assessment examples across elementary, middle, and high school.
The examples included:
Elementary: A first-grade playground design task that involved investigating light and materials.
Middle School: A musicians and hearing loss investigation focused on sound energy.
High School: The Ghost Cats investigation involving ecosystem interactions, population data, and energy flow.
Participants were assigned examples to examine and then shared their observations with the larger group. As someone who facilitates professional development, I appreciated how the presenter addressed different grade bands rather than focusing on just one level. The activities provided opportunities for participants to analyze tasks, discuss where student thinking was visible, and hear perspectives from other educators. I also liked the collaborative discussions. Participants were actively involved in examining instructional materials rather than simply listening to information. This was an engaging, collaborative approach to professional learning that made the content relevant to educators across grade levels.
My Takeaways from the Session
I left this session thinking about the relationship between AI, sensemaking, and rigorous assessment design. What made the session excellent was that it consistently returned to student thinking. Whether we were analyzing OpenSciEd student work, discussing uncertainty in the sensemaking process, examining the Ghost Cats investigation, or exploring AI tools, the focus remained on what students understand and how they demonstrate that understanding. I also appreciated how the professional development was facilitated. The presenter incorporated visual explanations, student work analysis, grade-band examples, collaborative discussions, and opportunities to examine rigorous assessment tasks. As a Biology teacher and professional development facilitator, I can see opportunities to apply several of these ideas in my classroom and through MosesSTEMinc. This session reinforced the importance of designing assessments that give students opportunities to investigate phenomena, analyze evidence, develop models, and explain their thinking.
My biggest takeaway is that AI and rigorous science assessment can work together when student thinking remains at the center of the learning experience.
As educators continue exploring AI in science classrooms, I think one question deserves our attention:
When students engage with AI, how are we making their scientific thinking visible?
References and Resources
Badrinarayan, A. (2025). Surfacing brilliance through meaningful science assessment: Shifting practice to reflect today's students and today's standards. Corwin.
Explore the book through Corwin
Agoos, S. (2026, November 6). Designing for sensemaking: Using AI to strengthen instruction and assessment in K–12 science, technology & engineering [Conference presentation]. Colorado Science Conference.
Google Classroom: Create and assign Gemini Gems
MosesSTEMinc L3C
Supporting Science Instruction, Curriculum, and Professional Learning
