Three ways alternative grading addresses the EDUCAUSE Horizon report
It's more than just addressing AI
As a newly-minted administrator directing a university Center for Teaching Excellence, I’m learning to look at alternative grading from a strategic perspective. It’s very different from the tactical perspective that I had when I was teaching classes, using specifications grading on a daily basis. I’m not thinking about how to design, implement, and manage an alternative grading system in one of my classes. Instead, I’m thinking about how to support faculty who are doing these things. And to do that effectively, I have to understand how alternative grading touches on key systemic issues that matter to chairs, deans, provosts, and others.
Often this systems-level thinking involves looking at university strategic planning, and I’m doing that. But recently, another document has played an outsized role in shaping high-level conversations about teaching and learning: The 2026 EDUCAUSE Horizon Report (Teaching and Learning Edition). I’ve only completed one week on the job at this point but this document has come up centrally in at least three different conversations. Today I want to share how I believe alternative grading addresses at least three major threads in this very important report.
What is the EDUCAUSE Horizon Report?
EDUCAUSE is an influential, respected nonprofit organization that advances higher education through the strategic use of information technology and data, particularly by tracking educational technology trends and their impact on college teaching and learning. Although it has a distinctly technological flavor to it, EDUCAUSE and its research are considered to be trustworthy and significant for higher education generally.
The annual Horizon Report examines key trends, emerging technologies, and innovative practices in higher education. While editions vary in scope—such as the 2025 “Data and Analytics” or 2024 “Cybersecurity and Privacy” editions—the “Teaching and Learning” edition specifically focuses on how technology impacts classroom instruction and course design.
Developed by panels of experts using the STEEP framework (Social, Technological, Economic, Environmental, and Policy lenses), the report outlines forward-looking pathways for higher education. This year’s report introduces a “Signals of Change” section to highlight early indicators and cutting-edge innovations shaping the future of higher education.
The 2026 Horizon Report listed fifteen different observations (three per category in the STEEP framework) about the future of higher education. Unsurprisingly, artificial intelligence plays a large role in this year’s report. At least four of the observations are directly related to AI — and three of those four are directly addressed by alternative grading. Let’s go into the details.
Alternative grading and trust in information
The report points out that AI is reshaping how humans trust in information and states:
“Colleges and universities will increasingly need to help students focus on the skills underneath any tool: evaluating claims, checking evidence, explaining reasoning, and verifying information. This requires clearer guidance on responsible AI use, as well as a shift in assessment models that value depth of understanding and thinking processes over polished output.”
It’s interesting that the report calls out “assessment models”, apart from any technological considerations. I believe it’s referring to approaches to assessments such as returning to in class paper exams or oral exams. Re-examining our models is long overdue and a good idea. But to achieve maximum impact in restoring trust in information, grading practices need re-examination as well.
How does an assessment model “value depth of understanding and thinking processes”? The way an assessment model values anything is by the way that the assignments in that model are graded. In a traditional points-based, one-and-done environment, what’s truly valued is the product, not the process. There are no feedback loops to speak of, so although a traditionally graded assessment might assign point values to the various skills that the report points out (evaluating claims, checking evidence, etc.), students don’t get practice in building those skills. They are merely audited on the completion of tasks that use them. This mitigates against growth.
In an alternative grading environment, however, there are many options that fit with the overall goal of growing students’ universal skills – and their ability to determine the value of information and produce reasoning that others can trust. For example, in writing a research paper, the different skills that are valuable to research pointed out by the report could be given as separate parts of an overall research project assignment. For example, students could have part of the research project in which they simply evaluate claims, another where they check evidence, another where they explain their reasoning, and so on. Each of these parts can be graded on specifications, or ungraded, with feedback loops in place for students to repeat attempts that don’t meet an appropriately high standard.
It’s the feedback loop part of this that makes alternative grading amenable to actually building the skills in the report. Having reattempts without penalty and helpful feedback incentivizes not skipping the hard work of making sense of information.
Alternative grading and relationships between instructors and students
The Horizon Report also points out that AI is reshaping relationships between students and instructors, and not in good ways. It points out that AI is becoming a primary form of academic support as students sometimes feel more comfortable seeking help from a chatbot than they do from their professors. The report points out that a misalignment and expectations on how to use AI between instructors and students, along with limited transparency, a growing use of janky AI detection tools and more is creating mutual distrust and suspicion.
But this is just exacerbating existing strained relationships between instructors and students that can be traced back to multiple sources, including the use of traditional grading practices. Many of our guest posts in the past that highlight instructors who changed from traditional to alternative grading practices point out that prior to their switch, they were seen as the gatekeeper of points and therefore of success, setting up unhealthy and sometimes adversarial relationships. But following the switch, students were more collaborative, more relaxed, and less stressed about the course, and therefore more likely to come to their instructor for help because the feedback loops that undergird the course make help-seeking behaviors pay off.
The Horizon Report states:
“Looking ahead, institutions will need clear, consistent norms for ethical AI use, along with a clear plan to invest in the human parts of higher education that AI cannot replace, including mentoring, judgment, and learning designs that keep trust, belonging, and student growth at the center.”
Apart from the call for norms for AI use, all of the items in the second half of that paragraph are fostered by alternative grading practices and diminished by traditional grading practices. Instructors working with students in an alternative grading setting where feedback loops are the main driver of activity and where reattempts without penalty and helpful feedback are available, is a radically human experience. Alternative grading practices build trust, fosters a sense of belonging, and drives student intellectual growth.
Alternative grading and redefining teaching and instructional design
Over the last couple of years, I’ve been blogging about how AI has forced me to redesign my assessment and grading practices and my classroom teaching because of concerns over academic dishonesty. Briefly and frankly, I came to the conclusion that there should be no asynchronous outside of class work in my class anymore because it’s too exposed to the risk of AI cheating.
The Horizon Report tends to agree with me on this by saying:
“As student AI use becomes routine, traditional assessments are being questioned for their ability to reliably measure learning, especially due to concerns over cheating. In response, instructors are facing pressure to redesign assessments to focus on more authentic demonstrations of learning that emphasize process, reasoning, collaboration, oral explanation, and critical evaluation of AI outputs, with some faculty even reverting to traditional methods such as Scantron forms, blue books, and in-person proctoring.”
I think it’s even worse than the horizon report makes it sound, because first of all, the idea of “measuring learning” is somewhat problematic, and also because instructors are not merely “facing pressure” to redesign assessments: They have to redesign many if not most of them, because traditional assessment practices were already broken well before AI showed up. The emergence of AI simply ends our ability to ignore it.
Traditional grading practices erode all of the authentic demonstrations of learning that the report points out, by placing arbitrary point values on each of these authentic areas and by creating one and done situations where only the product is valued, not the process. On the other hand, alternative grading practices, which are based on feedback loops and iteration until a standard is met that is appropriately high, foster growth in each of these areas. So as I noted above, it’s not enough just to change the assessment models. The grading processes that are used on those models must also be changed to something that is more natural to the human learning experience and therefore less susceptible to AI.
Proving the value of higher education
I was on a flight recently and sat next to a college student, a bright and mature Chemical Engineering major who was returning to an internship. When I asked him how well his coursework prepared him for his job, he didn’t seem to have much to say, other than to describe the grades he got. If I didn’t know his professional situation, I would have had a difficult time determining the value of his education — even though the metrics of the outputs (the grades) were clear.
The three issues I’ve described here are really about how alternative grading addresses AI’s impact on higher education, and therefore indirectly touches on how alt-grading can be good for higher education generally. But the overarching message of the Horizon Report is that higher education is facing pressure like never before to defend itself — to prove its value and improve its return on investment. Alternative grading can help here too. By reducing DFW rates, we can improve overall retention rates and lower the cost of degree completion. By having students engage in authentically growth-oriented classroom and assessment practices, student work can become more visible and more than just the grade attached to the work.
So as I think about how to encourage those “above” and “below” me in my position now – the administrative upper-class, and instructors, respectively – to support practices that address numerous key issues all at once, alternative grading is right at the top of the list.