This resource is provided by ACSA Partner4Purpose Learning Genie and was written by David Ross.
For decades, the architecture of school assessment has been built around a familiar set of indicators: grades, attendance, discipline referrals, and standardized test scores. District dashboards display these numbers with impressive precision. Leaders can track them by school, grade level, and subgroup.
The problem is not the data. The problem is timing. Most of the indicators we rely on are lagging signals. They confirm that a problem has already taken root. By the time a student’s grades fall or chronic absenteeism appears in a report, disengagement has often been present for months.
In an era shaped by generative AI — where information is abundant and content production increasingly automated — schools face a more urgent question: Are we measuring the capacities that actually matter for the future?
Agency-based learning begins with a simple premise. In a world where machines can generate essays, presentations, code, and analysis in seconds, the competitive advantage shifts away from producing information and toward directing learning itself. The distinguishing human capacity becomes the ability to set goals, identify pathways, monitor progress, and persist through difficulty.
If that is the case, then our assessment systems must evolve accordingly.
The Limits of Traditional Indicators
Traditional school metrics were built for a different educational era. In the industrial model of schooling, the central task was delivering content efficiently and verifying whether students had absorbed it. Grades, test scores, and seat time functioned as proxies for learning. They were imperfect, but they aligned reasonably well with a system focused on coverage and compliance.That alignment is now under pressure from two directions.
The first is student disengagement. Across the United States, districts continue to grapple with elevated levels of chronic absenteeism and declining engagement. Attendance policies can require presence, but they cannot generate investment.
The second is artificial intelligence. When generative AI tools can produce competent written products instantly, assignments that focus solely on the final artifact lose much of their diagnostic value. A polished essay or slide deck no longer guarantees deep understanding or ownership of the ideas it contains.
This shift forces educators to confront a deeper question: If the product can be automated, what should assessment actually measure? The answer increasingly points toward process and agency.
Agency as a Leading Indicator
Research in psychology and education offers a useful framework for thinking about this shift. One of the most relevant is Hope Theory, developed by psychologist Charles Snyder.
Hope, in this research tradition, is not an abstract emotion. It is a measurable cognitive framework composed of two components:
- Agency (willpower): the belief that one can initiate and sustain action toward a goal.
- Pathways (waypower): the ability to generate strategies for achieving that goal.
Together, these capacities form what researchers describe as goal-directed thinking. Students with strong agency believe their effort influences outcomes and can identify multiple routes toward success when obstacles arise.
Decades of research connect these constructs to academic persistence, engagement, and performance. In other words, the psychological architecture of hope functions as a predictive indicator of student success. Schools, however, rarely measure it. Instead, districts rely on downstream metrics (the aforementioned grades and attendance) that reveal disengagement only after it has become visible.
Agency-based learning suggests a different approach by treating agency as an upstream diagnostic.
From Philosophy to Measurement
Fortunately, measuring agency does not require elaborate instruments. Snyder and his colleagues developed a concise six-item Hope Scale for Children, which asks students to respond to statements such as “I can think of many ways to get the things in life that are most important to me,” or “I energetically pursue my goals.”
When administered consistently, these responses can be aggregated to generate a Student Agency Score. This metric translates psychological research into the operational language of district dashboards.
This score is not intended to label individual students. Rather, it functions as a system-level indicator that leaders can examine across several dimensions:
- Average agency scores by grade level or school
- Distribution bands showing the proportion of students with low, moderate, or high agency
- Changes over time following instructional or programmatic shifts
- Correlations with attendance, persistence, or course completion
Viewed this way, agency becomes less a philosophical aspiration and more a diagnostic lens. The goal is not surveillance. The goal is insight. If a ninth-grade cohort shows a sharp decline in agency scores between fall and spring, leaders gain an early signal that something in the learning environment is undermining students’ sense of progress and control. Interventions can begin before disengagement appears in attendance or grade reports.
Guardrails for Responsible Use
Any discussion of new assessment tools raises legitimate concerns about misuse. District leaders must approach agency metrics with clear ethical guardrails.
First, agency scores should never be used for teacher evaluation. The constructs involved are influenced by numerous contextual factors beyond the control of any individual educator. Second, data should remain aggregated and improvement-oriented, not attached to individual student labels. Third, districts should treat agency metrics as part of a broader continuous improvement cycle, combining them with qualitative observations, student feedback, and classroom artifacts.
When used responsibly, the goal is not ranking schools but strengthening learning environments.
What Agency Looks Like in Practice
Measurement alone does not create agency. It simply makes the invisible visible. The real work occurs in classrooms where teachers design experiences that allow students to practice goal-setting, strategic thinking, and reflection.
A striking example comes from a transitional kindergarten classroom in California. Allison White, a veteran teacher in the Los Altos School District, recently experimented with using an AI unit-planning tool to co-create a short inquiry unit with her students.
The idea was straightforward: allow the curiosity of five-year-olds to shape the curriculum while ensuring alignment with the California Transitional Kindergarten Learning Foundations.
The inquiry began when White noticed her students had developed a fascination with volcanoes. After reading several nonfiction books together, she asked a simple question: “What do you want to know?”
The class generated a list of questions, which some educators call “wonderings.” White entered these student-generated questions into an AI planning tool along with the domains she wanted to address (literacy, science, and visual arts). Within minutes, the system produced a three-week unit outline. Importantly, the AI did not determine the topic or the questions. The students did.
Over the following weeks, the classroom transformed into a small laboratory of volcanic exploration. Students constructed volcanoes in the sandbox, examined diagrams of magma movement, and incorporated new vocabulary into play and conversation. One child remarked during a discussion, “Lava moves slower than I thought.” Another used a hand gesture the class had invented to distinguish between magma and lava.
These moments may seem simple, but they reveal something profound. Even at five years old, students were exercising the core components of agency, including curiosity, hypothesis formation, experimentation, and reflection.
White later noted that engagement during the unit was noticeably higher. When students recognized that their questions had shaped the lessons, participation increased across the classroom.
AI, in this case, did not replace the teacher or dictate instruction. It acted as a structuring tool, allowing White to organize student inquiry quickly while maintaining standards alignment. The result was not scripted learning but amplified agency.
Assessment in the AI Era
Stories like this illustrate why assessment systems must evolve alongside instruction.
In an AI-saturated world, evaluating learning solely through finished products will become increasingly unreliable. Machines can generate competent artifacts. They cannot replace the human capacity to set goals, adapt strategies, and persist through uncertainty.
Those capacities must become visible in our assessment systems. Agency-based learning does not abandon accountability. It reorients it.
Instead of asking only “What did students produce?” we must increasingly ask:
- What goals did they set?
- What strategies did they try?
- How did they adjust when obstacles appeared?
- What evidence shows they believed their effort could change the outcome?
These are not soft skills. They are the cognitive infrastructure of learning itself.
A Different Kind of Dashboard
For district leaders, the implication is clear. The next generation of school dashboards should not focus exclusively on academic outputs. They should also illuminate the psychological drivers that make those outputs possible. Grades and test scores will always matter. But they should be complemented by indicators that reveal whether students believe they can chart a path toward success.
When agency becomes visible in system data, schools gain an early-warning signal for disengagement and a guidepost for improvement. In an educational era increasingly shaped by artificial intelligence, the item of scarcity is no longer information. The scarcity is human direction.
And the first step toward cultivating that capacity is learning how to measure it.
David Ross is a writer and consultant in the U.S. and China who focuses on implementing generative AI in K-12 settings. He is the retired CEO of the Partnership for 21st Century Learning and the former Senior Director for the Buck Institute for Education.
Bibliography
Anderson, M. (n.d.). The factory model of schooling. Leading Great Learning. https://leadinggreatlearning.com/the-factory-model-of-schooling/
California Department of Education. (2026). California Preschool/Transitional Kindergarten Learning Foundations (PTKLF). https://www.cde.ca.gov/sp/cd/re/psfoundations.asp
Edutopia. (2015, November 10). “I wonder” questions: Harnessing the power of inquiry. https://www.edutopia.org/practice/i-wonder-questions-harnessing-power-inquiry
Horowitz, J. (2018). Why it is better to lead than lag: Leading and lagging indicators for education. Institute for Evidence-Based Change. https://www.iebcnow.org/wp-content/uploads/2021/03/IEBC-Leading-and-Lagging-Indicators.pdf
Hrynowski, Z. (2024, August 21). K-12 schools struggle to engage Gen Z students. Gallup. https://news.gallup.com/poll/648896/schools-struggle-engage-gen-students.aspx
Malkus, N. (2025, June). Lingering absence in public schools: Tracking post-pandemic chronic absenteeism into 2024. American Enterprise Institute. https://www.aei.org/research-products/report/lingering-absence-in-public-schools-tracking-post-pandemic-chronic-absenteeism-into-2024/
McTighe, J., & Tucker, C. (2022). Developing self-directed learners by design. Educational Leadership. ASCD. https://www.ascd.org/el/articles/developing-self-directed-learners-by-design
Lopez, S. J., Rose, S., Robinson, C., Marques, S. C., & Pais-Ribeiro, J. (2009). Measuring and promoting hope in schoolchildren. Handbook of Positive Psychology in Schools.
Rand, K. L., Martin, A. D., & Shea, A. M. (2011). Hope, but not optimism, predicts academic performance of law students beyond previous academic achievement. Journal of Research in Personality, 45(6), 683–686. https://doi.org/10.1016/j.jrp.2011.08.004
Snyder, C. R. (1994). The psychology of hope: You can get there from here. Free Press.
Snyder, C. R. (2002). Hope theory: Rainbows in the mind. Psychological Inquiry, 13(4), 249–275.
