In online education, transparency is foundational to high-quality learning experiences (Ecampus Essentials, 2026). Transparency about who the instructor is, how the course set-up, what tools are available, and how learning will be assessed helps students trust the learning process. The emergence of generative artificial intelligence (AI) tools in higher education raises important questions about transparency in online courses, particularly regarding instructor and student disclosure of AI use.

Disclosure of AI use in online teaching and learning can be challenging for a number of reasons. First, it may be difficult to know what constitutes “use.” For example, is a Google search that resulted in an AI generated summary to gather information for course development considered “use”?  What about a student using generative AI tools to correct spelling, grammar or punctuation in a term paper? Second, it may also be difficult to assess what level of use warrants disclosure. Does using AI tools to generate ideas for the structure of a video lecture need to be disclosed, even if the content of the lecture was generated by the instructor without further AI assistance?

There are also real concerns about how disclosure might impact perceptions of the person using the tools and the quality of their work. Recent studies have found that in general, disclosing AI use across a range of tasks erodes trust (Schilke & Reimann, 2025). The “trust penalty” is particularly strong for women who disclose their use of generative AI tools in professional contexts; research has shown that women are disproportionately viewed as “lazier” and “less competent” (Chatoo, 2026; Gai, Hou, & Tu, 2025).

These research findings might explain why in a survey of faculty we conducted in the Ecampus Research Unit, 32% of online instructors reported never disclosing their AI use to students, even though 59% were using these tools for work (Aguiar et al., 2026). This stands in stark contrast to faculty expectations of their online students. About half of faculty respondents indicated that they allowed students to use AI tools in their online courses, provided that students properly disclosed their use.

This misalignment between faculty disclosure practices and expectations of students could disrupt the relational trust between instructor and student that is essential for high-quality online learning. Lack of disclosure may also result in additional work burdens for online students and faculty alike. Students may spend significantly more time documenting how they did not use AI to complete their assignments (Luo, 2024). And in our faculty survey, one respondent indicated that their grading time had increased by “300%.” Another instructor commented, “…it takes a lot of faculty time to grade AI outputs and/or screen them for falsified information” (Aguiar et al., 2025). According to this same instructor, the additional work burdens for online students and faculty creates a “lose-lose” situation.

While faculty disclosure of AI use clearly comes with some reputational risks, the long-term benefits likely outweigh current risks. When instructors model disclosure of AI use for students, it contributes to a culture of openness, honesty, and professional vulnerability that positively impacts trust between faculty and students in the online learning environment (Luo, 2024). This in turn leads to a higher quality learning experience for online students. And when faculty and students are open, honest, and transparent about their use of generative AI tools, this can also alleviate additional work burdens and subsequent feeling of anxiety and/or stress.

So, how can instructors model disclosure of AI use for students in online courses?  Based on our research with online students and faculty, here are some recommendations to get started:

  1. Define what AI use means– for you and your students. What tools are considered AI tools (e.g., Grammarly vs. ChatGPT), and what level of use is appropriate in a given context? This will help both you and your students clearly understand what use means and what may need to be disclosed.
  2. Clearly articulate when disclosure is required– for you and your students. Talk to colleagues about what use cases they think require disclosure. Provide students with specific examples from assignments in your courses to help reduce everyone’s feelings of uncertainty and stress.
  3. Provide specific examples of how to disclose. If you used AI tools to generate course materials, model disclosure by describing what tools you used and how you used them. A recent blog post by Ecampus instructional designer, Deb Mundorff, provides specific examples of how faculty have chosen to disclose their AI use with online students. If you did not use any generative AI tools, you can also let your students know in a disclosure statement (e.g., “No generative AI tools were used in the development of my course materials”).
  4. Openly discuss challenges and solutions related to disclosing AI use. Conversations about professional reputational risk are valuable to have with colleagues and with your students. These conversations can help build a culture of honesty and transparency in the use of AI tools.
  5. Iterate and revise disclosure guidelines and practices based on feedback from colleagues and students. AI tools and related applications are rapidly changing. Working together to reconsider approaches in real-time will help mitigate challenges and stressors as they arise. Inviting students to contribute to disclosure guidelines can inform current and future policies that work for both instructors and students.

Disclosing use of generative AI tools is an evolving challenge that can feel fraught and overwhelming. Disclosure may come with reputational risks that can disproportionately harm some individuals more than others. That is why building a culture of transparency with AI use is essential and something we can and should do together. It will require practice, patience, and iteration. Sometimes we will make mistakes, but what will matter most is our willingness to be open and honest about our AI use and support others to do the same. This supports trusting relationships that promote a healthy online teaching and learning environment.

Check out some additional resources from both Ecampus and the Center for Teaching and Learning; and reach out to share your thoughts with us. We’d love to hear your approach to disclosing AI use in your work.

References:

Aguiar, N. R., Coltharp, A., Dello Stritto, M. E., Jorgensen, J., Kronser, Z. E., Pearson, A., Piacenza, S., & Reese, D. (2026). Online Faculty Perceptions of Generative AI. Oregon State University Ecampus Research Unit. https://ecampus.oregonstate.edu/ research/publications/

codeforgoodnow. (2026). The AI Judgement Penalty. https://codeforgoodnow.com/wp-content/uploads/2026/06/Research-Report-The-AI-Judgment-Penalty.pdf

Ecampus Essentials. (2026, August 11). Policies & Standards- Oregon State University Ecampus. Retrieved August 2026 from https://ecampus.oregonstate.edu/faculty/standards-principles/ecampus-essentials/

Gai, P. J., Hou, J, & Tu, Y. (2025). Competence Penalty Is a Barrier to the Adoption of New Technology (May 11, 2025). Available at SSRN: https://ssrn.com/abstract=5255039 or http://dx.doi.org/10.2139/ssrn.5255039

Lou, J. (2025). How does GenAI affect trust in teacher-student relationships? Insights from students’ assessment experiences. Teaching in Higher Education, 30(4), 991-1006. https://doi.org/10.1080/13562517.2024.2341005

Mundorff, D. (2026, August 10). Whose Work Is It Anyway? Writing an AI Statement Worth Reading. Ecampus Course Development & Training Blog. https://ecampus.oregonstate.edu/faculty/inspire/2026/08/10/whose-work-is-it-anyway-writing-an-ai-statement-worth-reading/

Schilke, O. & Reimann, M. (2025). The transparency dilemma: How AI disclosure erodes trust. Organizational Behavior and Human Decision Processes, 188, 104405. https://doi.org/10.1016/j.obhdp.2025.104405.

AI disclosure statement: No generative AI tools were used in the development, writing, and review of this blog.

One instructor I work with said this recently:

I like that idea! So then students can start their project early on, get some classmates’ comments, and submit their final version

This response came after we spent some time discussing the challenge that comes with a term project that gets submitted as a one-time assignment at the end of the term.

Some instructors express interest and concerns about developing a term project and how to implement it in an asynchronous course. A project can be a great opportunity for engaging students with the content, learning from and with peers, and applying the concepts they learn in more practical and hands-on activities; and for it to be clear and well-implemented, it needs structure. However, it is not always clear for some instructors how the project could be spread out through the term in ways that allow students to move steadily and how to support them in each step of the project. Concerns about the implementation also relate about how to allow space for students to learn from one another through feedback that is effective. This blog describes how to structure and scaffold an individual project to offer students options to connect topics to their own interest and engage in peer feedback.

Project Overview and Goals

When I collaborate with instructors in developing this kind of assignment, we spend quite some time reviewing the project goal and details, brainstorming ideas for the milestones students will need to accomplish, the timeline, and the delivery of the final product.

The structure of the term may not always be evident from the start. A few questions that can guide the development of a term project that I ask instructors include:

  • What is the overall goal of the project?
  • Is this an individual or a collaborative activity?
  • What will students need to create at the end of the term (e.g., report, presentation, collage)?
  • How do students get started with it?
  • What are the opportunities through the term to seek feedback?
  • Are there resources needed to support this kind of activity?
  • How many stages does the project include?

The last question in the list above sometimes puzzles some instructors. The term project may not have been conceived as a set of multiple stages; instead, it is often a one-time, final assignment students need to turn in during finals week. Students may not get feedback or if they do, it won’t be actionable since there is no opportunity to revise and resubmit the project after they submit it.

In contrast, a multi-stage assignment may help students get the work done more steadily, in small portions, potentially helping them to manage the course load more effectively, and creating a project that shows how students apply what they learn. To respond to the need to reenvision the term project as a robust and structured activity that allows students to build their knowledge and skills step-by-step, instructors can resort the Transparent Assignment Design Framework< (TAD) and adapt it to design a term project. This framework is helpful in designing activities to promote student success and equity.

Adapted TAD for Term Projects

An adapted version of the TAD framework involves considerations such as structure (the core of TAD), scaffolding, resources, and evaluation. These considerations are described in the following sections.

Structure

Structure provides the connection and organization among all the elements of the project, from the goal to its deliverable. A clear project structure provides students with a better understanding of the value that the project brings to their learning, the stages it involves, the expectations to meet, the time and planning, and the resources needed to complete it. For setting up the structure of a project, you will need to perform two types of thinking:

  1. “think backwards” and determine how will students arrive to the final product and what resources or smaller tasks they will need to complete for that purpose. This design strategy can help you identify the number of parts or stages the project will need to include.
  2. “think laterally” and determine what content and activities in the term can support a particular stage or part of the project. This design strategy can help you balance the work for students (and for you as well when grading!) in each weekly module, align the project with the learning outcomes, and identify scaffolding mechanisms.

For example, the structure for a term project about theories of learning may include:

  1. Topic proposal
  2. Outline<
  3. Annotated bibliography
  4. Major scholarly works in conversations
  5. S*** First Draft (it comes from the work of Anne Lamott, Bird by Bird
  6. Peer review
  7. Revised draft/final revised version

Scaffolding

Scaffolding is the support and guidance that students received along the project tasks or stages. This mechanism allows students to complete more manageable pieces without getting confused or overwhelmed. Scaffolding can be provided through a set of tasks, from simple to more complex, sequencing them more cohesively. Through scaffolding, students may feel confident of their learning process and eventually complete tasks more independently. In the example above, the scaffolding for the proposal (step 1) and the annotated bibliography (step 3) could include the following guidance and support:

  • Step 1: Topic proposal:
    • Provide students with a list of broader ideas to get started or a list of topics that meet the expectations and are aligned with the outcomes
    • In addition, offer students an option where they select a topic on their own and seek instructor feedback/approval.
  • Step 3: Annotated bibliography
    • Provide guidance on the scholarly sources, format, and citation style
    • In addition, give students an example as a point of reference

Additionally, consider the content and other course activities that can support the project (lateral thinking). For example, ask yourself:

  • What are other activities/assignments in the week/module?
  • Do any of these activities relate closely to a project stage? If so, how that activity could provide students with a thinking exercise that allows them to connect the concepts in the project? (e.g., a discussion activity could engage students in exploring multiple perspectives, resource sharing, or idea exploration that they can incorporate in the outline or first draft of the project)
  • Do any of the course activities overlap with a project stage? If so, can they be merged and deepened to help students meet the intended outcomes?

Evaluation

With any kind of graded activity, it is also important to provide students with clear information about how their project and its stages will be evaluated. The criteria for evaluating the project will help students understand more clearly the expectations and their level of quality. These criteria can be provided at each stage of the project, if each stage will be evaluated or graded separately. For the example above, the evaluation can include:

  • Checklist for expectations
  • Example of a good quality project
  • Grading criteria (a rubric or guidelines that describe performance levels)

In short, when implementing a project in your course, think about it as a pleasant, fun journey rather than a challenging odyssey filled with hardships where students struggle trying to figure out what to do next.

Oh…you may be wondering about the instructor and their project idea that I mentioned at the beginning of this blog? Well, the project is well-structured with several small, manageable tasks (e.g., discussions, individual assignments, peer review) each week that help students build the project, one step at a time. The instructor is satisfied with the project structure and is eager to get that implemented in the course next term!

Do you have a project idea for your course? How are you thinking about its structure and deliverables? What kinds of guidance are you considering including? I would love to hear what ideas you have!

AI disclosure: No AI tool was used for the planning, drafting, or revisions of this blog.

Last year, as the instructional designer for a software engineering course, I knew the instructor had leaned on AI throughout the course’s development. I suggested he be transparent about it with his students. After all, he was asking the same of them.

What he wrote back went further than I expected. His syllabus statement opens not with rules for students but with an accounting of his own use. He names Claude the second author of his course, separates his own contribution from what emerged from that collaboration, and only then turns to what students may do. After getting his permission, I shared it with my team. Within minutes, colleagues were asking to keep it as an example. One had just come from a webinar where participants argued faculty shouldn’t have to disclose their own AI use at all.

Their reaction was telling. Instructional designers see a lot of syllabus statements, and this one stopped people. It fit what I keep hearing from faculty: they are no longer asking whether to address AI in the syllabus, but how to do it well.

Plenty of instructors still open with a different question: can I just ban it? I understand the impulse, but it is the wrong place to start. AI is already in your course, in your students’ hands, whether or not your syllabus mentions it. When Bowen and Watson (2026) talked with students, some said that if an instructor never mentioned AI, they took the silence as a “don’t ask, don’t tell” policy. Students brainstorm with it, run drafts past it, and ask it to re-explain what you covered last week. Ban it outright, and you’ve got a policy you can’t enforce and students who use it quietly anyway. I made that case at more length in Students Don’t Want AI Bans, They Want a Seat at the Table. What your syllabus has to sort out is where AI helps students learn and where it gets in the way, and then say which is which in plain terms.

What changed is not that AI arrived. It is what AI can do.

A couple of years ago, using AI meant asking a chatbot for a paragraph and pasting the result. That is still common. What is different now is scope. Generative AI answers a prompt. Agentic AI takes a goal and works through the steps to reach it. Point it at an assignment, and it can plan, draft, and revise the work. It hands back something finished, with little from the student.

This shift matters for your syllabus, because it changes what doing the work means. A policy written for a student who copies one answer from an AI tool is aimed at the wrong target. That does not mean everything is permitted; it means deciding on purpose which uses belong in your course and which do not, rather than letting silence decide for you. Here is the question your syllabus has to answer: in this course, what part of the work has to be the student’s own, and how will they show it?

A good statement does two jobs.

The first is obvious. Tell students what they may and may not do with AI. Do it assignment by assignment, since a tool that undermines one task may be right for another. Tie the rule to the point of the assignment. “Do not use AI to draft your reflection, because the thinking is the assignment” gives students a reason they can act on. A blanket ban does not.

The second is easier to forget. Say how you use AI too. If you used it to build a study guide, draft feedback, or write practice questions, tell students. It models the honest, purposeful use you are asking of them. At OSU, Ecampus collects AI tools and faculty support resources, including sample syllabus and assignment statements you can adapt.

Write it like you mean it.

The statements that work read like a person wrote them, not a policy office. A good one names what students can do that a model cannot, and makes building those skills the point of the course. Remind students that their own thinking and voice are worth developing; that does more to shape how they work than any list of don’ts. And if your campus hands you boilerplate policy language, treat it as a floor rather than the finished statement, then build your own on top.

The software engineering instructor I opened with is the clearest example I have. His statement spends several paragraphs on his own use of AI before it says a word about students’. He marks off what is his: the structure, the pedagogy, the assessments, and the running software example he wrote for his own textbook. Then he names three ways he worked with AI: as a conversational partner, as a second author on the course materials, and as a software developer. He is specific about the tool and the timeline, down to which modules he wrote before starting to use it. And he is clear that the collaboration made the course better, more engaging and more technically sound, rather than easier.

“I want to be completely transparent about how I have used GenAI to author this course.”

Only then does he turn to students:

“I strongly encourage you to use GenAI in this course as thoughtfully, conscientiously, and deliberately as I have used it to develop this course.”

Chris Hundhausen, software engineering. Shared with permission.

That last line works because of everything before it. He earned the right to ask by putting his own record on the table.

Whatever you write, plan to revisit it. The tools will change before next term. So will the way your students use them. A statement you can update in ten minutes each term beats a perfect one you write once and forget.

Offer a seat, not a rulebook.

So decide which part of the work must be your students’ own, and how they will demonstrate it. Be clear with them about what learning in your course requires. You can ask your students to help, since they are much more likely to honor and understand a policy they helped build.

A note on process: the argument, examples, and choices here are mine. I used an AI assistant to pressure-test the structure and tighten the wording, the same transparency this piece asks of you.

Reference

Bowen, J. A., & Watson, C. E. (2026). Teaching with AI: A practical guide to a new era of human learning (2nd ed.). Johns Hopkins University Press.

Authors: Mary Ellen Dello Stritto & Dana Simionescu

What We Did and Why

When we design online courses, we make hundreds of small decisions about where to put things or how to introduce them. Should assignment instructions come before or after the readings? Does anyone look at the Module Learning Outcomes? Most of these calls get made on how we know people learn, what we think learners do, or what the program has always done, but rarely on concrete evidence about how students actually move through a Canvas course.

So, we asked them. During Winter 2026, we surveyed 463 Oregon State University Ecampus students about how they navigate Canvas, where they look for information, how they engage with course elements, and what they wish we’d do differently. Respondents were experienced online learners, so the patterns below reflect habits built up across various courses and instructors.

Some results confirmed what we suspected. Others were genuinely surprising. Below are the takeaways we think are most useful for instructors and instructional designers.

Takeaways

  1. Consistency across courses is what students wanted most.

When we asked students about improvements they wished for, the most common answer was about consistency. Students described spending the first weeks of every term re-learning each instructor’s idiosyncratic Canvas setup rather than focusing on the course content itself.

This is a hard problem because instructor autonomy is real and valuable, but it points to a clear opportunity: shared design elements and practices within a program or department go a long way. Even standardizing a few things—where the syllabus lives, a common module structure, consistent naming for module pages—reduces the cognitive load students carry from course to course. This is a strong argument for using course design templates.

  1. For most students, the Modules page was the front door.

The clearest finding from the survey: most students lived in the Modules page. When asked where they go to find assignment information, 55% said the Modules page. When asked what they do first when starting a new module, 72% said they go to the Modules page and click the first item.

But it’s worth noting that a substantial minority of students navigated primarily via the To-Do list, the Canvas Calendar, the Grades page, or other tools. This is often because they’re managing a heavy course load and prioritize by due date across all their courses rather than working through one module at a time. The Modules page is the dominant pathway, but it’s not the only one, and a well-designed course needs to support both behaviors.

This suggests a few practical instructional design moves:

  • Make the Modules page complete and self-contained. Anything important (instructions, rubrics, readings) should be reachable from a module. If it isn’t in (or linked from) Modules, most students won’t find it.
  • Set module requirements for key pages. Canvas allows you to require students to “View” or “Mark as done” specific items in a module. This helps ensure students see important pages they might otherwise skip, and it activates the progress indicators on the Modules page—which several students specifically asked for, as a way to track what they’ve completed.
  • Sequence matters. About 71% of students report always or often going through module items in the order presented, so the order you set is the order most students will follow.
  1. The Calendar and To-Do list are critical infrastructure. Make sure everything is there.

When students manage workload across multiple courses, they rely on Canvas’s cross-course tools: the Calendar, the To-Do list, and the Grades page, which several students described as the most reliable place to see all assignments listed consistently.

The most actionable finding here is about due dates. Students reported that items are sometimes missing from the Calendar and To-Do list, which leads directly to missed deadlines. The single most-mentioned problem was discussion posts with two due dates (initial post + replies). When only one of the dates shows up in the To-Do list, students might miss the other.

A few concrete design ideas:

  • Make sure every graded item has a due date entered in Canvas. This is what populates the Calendar and To-Do list for the students who navigate by deadline rather than by module sequence. This includes graded items that are completed in an external platform.
  • Consider adding “to-do dates” to non-assignment pages like learning materials or assignment overview pages. Canvas lets you assign a to-do date to a page so it appears in students’ To-Do lists alongside graded items. This brings important non-graded content into the deadline-driven navigation paths some students rely on.
  • For discussions with two post deadlines, use Canvas’s new discussion checkpoints feature, which lets you set separate due dates for initial posts and reply posts. They’ll appear as distinct items in students’ To-Do lists and Calendars, which is exactly what students in the survey were asking for.

Bar chart titled “Where do you most often go for assignment information? (n=458)”. Responses: Modules 55%; To-do list 13%; Canvas Calendar 13%; Assignments link 9%; Other 5%; Syllabus page 4%. For "Other", the most frequent is: Grades.

  1. What students actually read (and what they didn’t).

Of all the course elements we asked about (announcements, rubrics, assignment purpose statements, and module learning outcomes), rubrics came out on top across every measure: 80% of students read them always or often; 89% rated them useful or very useful; 55% checked the rubric specifically when looking for feedback on a graded assignment.

By contrast, module learning outcomes had the lowest engagement of any element we asked about. Only 45% of students read them always or often, and only 45% found them useful. About 15% rated them not at all useful. It might be worth experimenting with how outcomes are presented, where they appear, or how they’re connected to specific activities.

While announcements are the most consistently read course element in the survey (78% always or often), about 22% of students read them only sometimes or never, a significant subset of students. This suggests you should not put essential, course-critical information only in an announcement. Deadline changes, an assignment clarification, a required reading update— these need to exist somewhere durable in the course such as the module itself, the assignment page, or the syllabus in addition to any announcement you send.

  1. Students often previewed assignments before doing the readings.

About 75% of students indicated they always or often complete learning materials before starting an assignment, which is reassuring. But the open-ended responses complicated the picture: many students described previewing the assignment first, then working through the readings and videos with a clearer sense of what to focus on.

This is a reasonable strategy, and the design implication is about making the connection between learning materials and assignments explicit. When students preview an assignment, you want them to be able to tell, quickly, which readings and videos actually prepare them for it. Call that connection out directly on the assignment page, in the learning materials, or both! That will help students study with purpose.

Semi-donut chart titled “How often do you view learning materials before starting assignments? (n=458)”. Responses: Often 46%, Always 29%, Sometimes 24%, and Never 1%.

  1. Course introduction materials were valued and used as ongoing reference.

In our Ecampus courses at Oregon State University, the introductory section is called the “Start Here” module, and nearly all students (94%) engaged with it. The specific format will vary by institution and program, but the underlying behavior is informative regardless of what you call your opening module.

A couple of patterns stood out:

  • Students returned to introductory materials throughout the term, not just at the start. They come back to look up course-specific information they need: the syllabus, grading policies, instructor contact info, the schedule, materials lists. Whatever you put in your opening module, students are treating it as a reference document rather than a one-time orientation.
  • Students valued course-specific information, rather than repeated institutional or platform content. Several experienced students described feeling that intro materials contain too much familiar content such as generic LMS tutorials, university-wide policies, boilerplate language that doesn’t change from course to course. They indicated that the genuinely useful course-specific information sometimes gets buried.

The takeaway here is that the introductory part of a course functions as a course-specific reference, so the design move is to make the truly course-specific content (your syllabus, schedule, assignment overview, contact info, policies) easy to find and return to—and to be thoughtful about how much repeated cross-course content shares that space.

  1. Mobile is for monitoring, not learning.

Our results showed that mobile usage was widespread but specific in purpose: 78% of students checked grades on their phones, 70% checked due dates, and 62% read announcements. But only 8% took quizzes, 8% completed assignments, and 13% participated in discussions on mobile devices.

The pattern is clear: students used phones to monitor and plan and they used computers to actually do the work. These results suggest you probably don’t need to redesign assignments to be mobile-first, but the quick-check pathways (grades, due dates, announcements) should be especially clean and reliable.

  1. Feedback is sought but hard to find consistently.

94% of students checked the gradebook Comments box for feedback. Roughly 59% also looked at annotated comments on the assignment, and 55% checked the rubric. Very few students skipped feedback entirely.

They told us the challenge is location. Students described frustration that feedback is input in different places depending on the instructor and the course. One student indicated they didn’t even realize that annotated comments existed.

If you use annotations or rubric-based feedback, tell students explicitly where to look for that feedback, ideally early in the term and again when the first graded assignment returned. Consider a dedicated section in your starting module, and/or a one-sentence announcement after the first round of grading such as “Your feedback is in three places: the Comments box, annotations on your submission, and the rubric.” This can save students from missing substantial parts of instructor feedback.

  1. Keep modules unlocked when you can.

Several students specifically asked for modules to remain unlocked so they could work ahead when they knew a busy week was coming. Locked modules were described as a barrier to time management, especially for students juggling multiple courses, work, and other obligations.

There are legitimate pedagogical reasons to release content sequentially: scaffolding, preventing students from getting overwhelmed, keeping a cohort moving together for discussion-based learning. But if you’re locking modules by default rather than for a specific reason, it’s worth reconsidering. Unlocked modules let the students who want to plan ahead do so without preventing anyone else from working through the course at the standard weekly pace.

Conclusion

If we had to compress the survey into one sentence for course designers, it would be this: students want predictable, modules-centered courses where everything they need is reachable from one place, due dates are accurate and consistent, and feedback is easy to find.

If you’re doing similar work at your institution or experimenting with any of these changes in your own courses, we’d love to hear about it.

“Discussions are boring”, “it is so hard for students to truly engage,” “the linearity of the discussions doesn’t help to navigate and find the threads”, “posts could be AI-generated.” If you’ve voiced one of these about discussion boards, you’re not alone. Oftentimes, I hear instructors express these concerns. Discussion boards were once considered spaces for bringing students together to engage in authentic and genuine conversations. However, in today’s AI world, the concerns are compounded by the risk that discussion posts could be generated by AI, lacking students’ own voices and personal investment in building community.

What is not working with discussion boards?

Discussion boards have been the default tool in online courses to create a space for students and their instructors to build community, engage in conversations, develop a sense of belonging, and learn from one another. The discussion board is the online counterpart of the physical classroom space.

However, over the years, it seems discussion boards are limited in that they hardly promote student engagement and expand conversations. For many instructors, discussions have become a formulaic “post one and reply twice” approach that lacks the features to truly motivate students to participate beyond that formula. It has even become more challenging to create an engaging asynchronous discussion in the AI era, as some students might feel tempted to use AI to generate posts.

What can work better?

Designing discussions that are meaningful and engaging may require evaluating their use and structure. The design of discussions sets the foundation for students to be in community, understand the value of the activity, and actively engage with others. This foundation should respond to a clear structure where the purpose, tasks, and criteria for success are clear.

There are multiple ways in which discussion boards go beyond that formula, becoming spaces that hold the community of learners together. It is possible to design and facilitate online discussions that are engaging and meaningful, and that serve as dynamic learning spaces where students get to know their peers, engage in community, and build trusting relationships with each other and with their instructor.

How to redesign discussions in the AI era?

An AI policy should be clearly articulated so that students know whether AI can be used or not in their discussion posts and replies. Instructors could opt to co-create an AI policy and involve students in identifying actions and tasks that could benefit from, or not benefit from, the use of AI in their discussions (and other assignments). However, given that discussions are one of the few course activities where the process of student-to-student interaction matters more than a polished product, an AI policy alone may not help address the concern that posts could be entirely generated by AI.

In response to the AI era, it becomes even more important to ensure that the discussion board has a clear purpose and meaning for students. A purpose statement highlights the importance, relevance, and value of students’ own work and voices, and shows students that their instructors are eager to read their contributions. AI cannot replicate students’ own thinking and authentic voices. A clear description in the purpose statement and instructions that repositions the value of authentic student voice and contributions would make it visible to students that their instructor is interested in what they bring from their own context, lived experiences, or observations.

To make discussion prompts more engaging and tempting for students to bring their own voices, ask students to bring an artifact such as a news item, a course reading annotation, or a real-world example to share with peers. Then, ask them to respond to peers’ artifacts rather than to peers’ opinions. The input is the student’s own context, which sidesteps the AI temptation and makes the engagement more natural, relevant, and genuine.

Don’t feel discouraged by the shortcomings of discussion boards; these are great tools for creating the social interactions and community so desired in an online environment, even in the AI era and its challenges for teaching and learning. The key to using discussion boards is to rethink them as spaces for genuine engagement rather than as superficial, perfunctory tasks. Rethinking and redesigning discussions could start with small steps, for example, if you do nothing else:

  • Articulate clearly why students need to contribute and participate in the discussion. What is the value they will get from being in the conversation and community with others?
  • Describe your AI expectations in the discussion instructions, not just link to or add them in the syllabus. How would students benefit from using or not using AI when posting and replying?
  • Make one discussion prompt more meaningful, connected to students’ lived experiences. How do the topics discussed relate to students’ lives and communities?
  • Add reflection questions to the factual/descriptive prompts. How do students’ thoughts evolve as they engage in the discussion? How would students apply the concepts in other scenarios?

If you have tried something else or different, I would like to hear about it.

The Skill Students Resist, and Employers Expect

I have a confession. When I was earning my M.S.Ed. online, I dreaded seeing “group project” on a syllabus. I juggled coursework with a full-time job, logged in at odd hours, and did most of the work on every collaborative assignment. My standards were higher, my contributions more substantial, my late nights longer. I carried the team.

Here’s the awkward part: so did everyone else.

Researchers have been studying this phenomenon since 1979, when psychologists Michael Ross and Fiore Sicoly found that group members consistently overestimate their own contributions (Ross & Sicoly, 1979). When you ask everyone on a team to estimate their percentage of the work, the numbers don’t add up to 100%; they add up to 114% on six-person teams. In studies of scientific coauthors, that number balloons to 167%. The bigger the team, the worse it gets: eight-person groups claim a collective 140% of the credit (Schroeder et al., 2016).

So if you or your students have ever finished a group project thinking I’m the only one who pulled my weight, congratulations. You’re not a uniquely burdened hero. You’re experiencing a well-documented cognitive bias known as egocentric overclaiming. And so is every other person on the team.

If you’ve ever been on the receiving end of student complaints about group work, this bias is a big part of why. Each student genuinely remembers their own effort more vividly than anyone else’s. The frustration is real, even when the workload was more balanced than it felt.

Why Collaboration Belongs in Online Course Design

That frustration is one reason faculty hesitate to assign collaborative work, and one reason students dread it. When the experience feels unfair (even when it isn’t), it’s hard to see the value. And yet, employers keep saying collaboration is exactly what they need. NACE consistently ranks teamwork among the top competencies employers seek (NACE Job Outlook, 2025). And with the rise of remote and hybrid work, the ability to collaborate asynchronously with a distributed team isn’t a nice-to-have; it’s the job.

Here’s what I find most compelling, and what I didn’t appreciate until years after finishing my own degree: the asynchronous collaboration I did as an online student may have been closer to real workplace collaboration than any in-person group project. Navigating shared documents, negotiating timelines with classmates in different time zones, and giving written feedback to peers I rarely interact with face-to-face, I was rehearsing exactly the skills I now use every day in a distributed professional team. I just didn’t know it at the time.

And the data supports this. Research published in the Journal of Asynchronous Learning Networks found that online programs emphasizing student engagement and collaborative practices achieved completion rates of 85% or higher, meeting or exceeding those of their face-to-face counterparts. Faculty and instructional designers are paying attention, and so are employers. The question, then, isn’t whether to include collaboration in online courses. It’s about designing it so students experience it as preparation rather than punishment.

Making It Work: Practical Strategies for Course Design

None of this means we should assign a group project and hope for the best. When collaboration is poorly designed with vague instructions, uneven accountability, and misaligned outcomes, student frustration is justified. Here are approaches that work across disciplines:

Start with community, not content. Introductory discussions can feel formulaic, “share your name, major, and a fun fact”, but they don’t have to be. When students share their backgrounds, goals, and even their anxieties about the course in week one, it lays the groundwork for everything collaborative that follows. The shift is subtle but real: from “strangers assigned to a group” to “people who know something about each other.”

Be intentional about team formation. Random group assignment is easy, but a little structure goes a long way. Consider using surveys to match students by schedule availability, working style, or shared interests. Some institutions use dedicated tools , at OSU, for example, Ecampus offers a group finder tool , while others have students use AI to synthesize individual preferences into a working team charter. The goal is the same regardless of approach: help students start from common ground rather than cold introductions.

Make contributions visible. Remember that overclaiming bias? One of the best ways to counteract it is to build in structured peer evaluation of team contributions, not a review of each other’s papers, but an honest assessment of how each member showed up for the group. When teammates rate each other on dimensions like participation, communication, and follow-through, it creates accountability and gives instructors insight they wouldn’t otherwise have. At OSU, Ecampus has developed a custom tool for this purpose. CATME is a widely used option, and even a well-designed Google Form with clear criteria can do the job. However you implement it, the point is to give students a structured way to reflect on, and be accountable for, their collaboration.

Design for real-world parallels. In the sciences, interdisciplinary teams can tackle complex problems that mirror real-world research collaborations by designing solutions, analyzing data, and presenting findings as a group. In business courses, teams can build a business plan or marketing strategy together, or work through a case competition where they analyze a real scenario and present recommendations. In the humanities, students can collaborate on oral history projects or produce a podcast episode together, work that requires negotiation, shared decision-making, and a tangible product. The key is making the collaboration itself part of the learning outcome, not just a delivery mechanism.

A Better Way to Think About It

The next time a student groans about a group project or a colleague pushes back on interaction requirements, it might help to share the research on overclaiming. Not to dismiss their frustration, but to reframe it: the discomfort of collaboration isn’t a design flaw. It’s the learning happening.

And if they still insist they did most of the work? Well. So did everyone else.


References

  • Herz, N., Dan, O., Censor, N., & Bar-Haim, Y. (2020). Authors overestimate their contribution to scientific work, demonstrating a strong bias. Proceedings of the National Academy of Sciences, 117(12), 6282–6285.
  • Moore, J. C., & Fetzner, M. J. (2009). The road to retention: A closer look at institutions that achieve high course completion rates. Journal of Asynchronous Learning Networks, 13(3), 3–22.
  • National Association of Colleges and Employers. (2025). Job outlook 2025. https://www.naceweb.org/career-readiness/competencies/
  • Ross, M., & Sicoly, F. (1979). Egocentric biases in availability and attribution. Journal of Personality and Social Psychology, 37, 322–336.
  • Schroeder, J., Caruso, E. M., & Epley, N. (2016). Many hands make overlooked work: Over-claiming of responsibility increases with group size. Journal of Experimental Psychology: Applied, 22(2), 238–246. See also: Why teams overinflate their contributions. Chicago Booth Review. https://www.chicagobooth.edu/review/why-teams-overinflate-their-contributions

The golden rule of link accessibility: links should be descriptive! For foundational information on the why and the how, see OSU Digital Accessibility – Links.)  Let’s dig deeper into a few common questions:

Can I use “click here” or “this” for my link text?

This practice is not ideal, and it’s best to avoid it. While WCAG does permit it when surrounding context provides enough information, you would not be creating a good experience for your audience. That type of text is not descriptive enough to show the user where the link will go, and it’s especially problematic if this text appears multiple times! Think of people skimming the content – whether visually or via assistive technologies. It’s much more helpful when the text clearly conveys the link’s function or destination. See an example below.

Side-by-side comparison of unhelpful versus descriptive link text. On the left, three sentences each use 'click here' as the link text. Below, a simulated screen reader link list shows three identical 'click here' entries with no way to distinguish them. On the right, the same three sentences use descriptive link text: 'OSU Digital Accessibility guidelines,' 'W3C WAI Functional Images,' and 'APA citation formatting examples.' The screen reader link list below shows all three distinct, meaningful labels.

Yes, you can use an image directly as a link or button. But! If the image serves as a link on its own, make sure to write alt text that describes the action initiated by the link. The example image below is linked to an interactive lesson about cat behavior. Therefore, you would use the alt text “Cat Behavior Interactive Lesson”, NOT describe the image. See more explanations and examples on the W3C WAI Functional Images page.

Example image for cat behavior lesson.

Citation styles may be strict, but they do allow some flexibility for online-only resources and materials outside of formal papers. The recommended practice is to link the work title and ditch the DOI or URL, like in the example below. Check out more examples and explanations for APA and for MLA.

Side-by-side comparison of two APA citations for the same source. On the left, the citation ends with a long DOI URL displayed as a hyperlink. On the right, the article title is hyperlinked instead and the URL is removed entirely, resulting in a cleaner, more accessible reference.

Try not to! Having redundant links may increase extraneous cognitive load, since people may wonder whether they go to the same place or need to click the link again. The article The Same Link Twice on the Same Page: Do Duplicates Help or Hurt? gives a detailed explanation of why this may cause problems. The Office for Digital Accessibility at the University of Minnesota also includes “Avoid Repetitious Links” in their Dos and Don’ts.

You may have noticed, on occasion, “ghost links” in Canvas. The link validator or accessibility checker says there’s a broken or duplicate link, but when you look at the text, there’s nothing there. However, if you switch to the HTML editor, you’ll find the link lurking underneath. In the example below, you can see that there are actually two links instead of one: the Assignment 1 link was not completely deleted when I replaced it with Assignment 2.

What happens is that sometimes, if you delete text without unlinking first, the link may persist. To avoid this situation, make sure to remove the links before deleting or pasting in text.

Rich text editor showing one link vs html editor showing two links.

BONUS link-related tip: Don’t underline regular text

Usually, links are underlined, and most people think of links when they see underlined text. This may be confusing when they try to access the link and it doesn’t work. In addition, underlining is just not a good way of highlighting information. For more information, see an article and video from Boise State University: Underlined text.

A sample paragraph where the phrase 'rubric and grading criteria' is underlined but is not a link, while 'Course Resources page' is an actual hyperlink.

These practices make your course more readable, easy to navigate, and overall, more enjoyable for your students!

This is it, the week you’ve been waiting for!

Open Education Week is an annual celebration that raises awareness of global efforts to make learning more “open” — that is, more affordable and accessible to students everywhere. Every March, this weeklong, online event gives educators and students an opportunity to learn more about open educational practices and be inspired by the work being developed around the world, including by Oregon State University faculty.

Please join us for Open Education Week this year to learn how you can get involved and make a meaningful difference in the lives of OSU students.

  • What: Open Education Week
  • When: March 2-6, 2026
  • Where: Fully online
  • Who: Higher education faculty, students and thought leaders

This blog post is a continuation from “Refining Rubrics & Assessments: AI as Design Support – Part 1“.

Using AI to Refine Rubric Language

In the previous post, I gave an assignment prompt to Copilot (as that’s the recommended tool at Oregon State University) and asked it to complete the task. For reference, here is the task.

Rubrics are often the weakest link in assessment design, particularly when descriptors rely on vague phrases like “meets expectations” or “demonstrates understanding.” One way to evaluate rubric clarity is to ask AI to self-assess its own response using the rubric criteria.

If the model can plausibly justify a high score despite shallow reasoning or inconsistent logic, the rubric may not be clearly distinguishing levels of performance. More precise rubrics specify what evidence matters and how quality differs, emphasizing reasoning, coherence, and alignment with course concepts rather than polish or length. Clear criteria benefit students, but they also make it harder for superficially strong work to masquerade as deep learning.


Rubric Analysis Prompt (Click to expand)

You are now acting as an external assessment reviewer, not a student.
You will be given:

  1. An assignment prompt
  2. A grading rubric
  3. A model-generated student submission (your own prior response)

Your task is not to grade the submission.
Instead, critically evaluate the rubric itself by answering the following:

  1. Rubric Vulnerabilities
    • Identify specific rubric criteria or descriptors that allow a high score to be justified through fluent but shallow reasoning.
    • For each vulnerability, explain what kind of weak or superficial evidence could still plausibly receive a high score under the current wording.
  2. Distinguishing Performance Levels
    • For at least three rubric categories, explain why the difference between “Excellent” and “Good” (or “Good” and “Satisfactory”) may be ambiguous in practice.
    • Describe what concrete evidence a human grader would need to reliably distinguish between those levels.
  3. AI Self-Assessment Stress Test
    • Using your own generated submission as an example, explain how it could convincingly argue for a high score even if underlying understanding were limited.
    • Point to specific rubric language that enables this justification.
  4. Rubric Strengthening Recommendations
    • Propose revised rubric language that makes expectations more explicit and evidence-based.
    • Emphasize observable reasoning, causal explanation, constraint awareness, or conceptual boundaries rather than general phrases such as “demonstrates understanding” or “well-justified.”

Constraints:

  • Do not rewrite the assignment prompt.
  • Do not assume access to course-specific lectures or materials.

Focus on how the rubric functions as an assessment instrument, not on pedagogy or student motivation.

Tone:
Analytical, critical, and concrete. Avoid generic advice.



You could use this directly by attaching a rubric, assessment prompt, and “submission”, or modifying it to your own situation.

Here is a section of the results it gave, along with the “thinking” section expanded to see the process of the generated answer:


(Copilot gave me an enormous amount of feedback, as expected because the rubric included a lot of generic language.)


Rethinking “Higher-Order Thinking” in an AI-Rich Environment

Frameworks like Bloom’s Taxonomy remain useful, but AI complicates the assumption that higher-order tasks are automatically more resistant to outsourcing. AI can analyze, evaluate, and even create convincing responses if prompts are static and unconstrained.

What remains more difficult to outsource is judgment. Assignments that require students to choose among approaches, justify those choices, identify uncertainty, or explain when a method would fail tend to surface understanding more reliably than tasks that simply ask for analysis or synthesis. When reviewing AI-generated responses, a helpful question is: What would a human need to know to trust this answer? Designing assessments around that question shifts the focus from output to accountability.

Instructors can strengthen authenticity by introducing under specified scenarios, realistic limitations, or prompts that require students to articulate how they would evaluate the reliability of their own results. These design choices don’t prevent AI use, but they make it harder to succeed without understanding when and why an answer might be wrong.


An Iterative Design Loop for Assessments and Rubrics

Using AI as an assessment design diagnostic and refinement tool can work best as an iterative process. Draft the assignment and rubric, test them with AI, analyze how success is achieved, and revise accordingly. The goal is not to reach a point where AI “fails,” but rather a point where success requires engagement with disciplinary concepts and reasoning. This mirrors quality-assurance practices in other domains: catching misalignment early, refining specifications, and retesting until the design reliably produces the intended outcome. Importantly, this loop should be finite and purposeful, not an endless escalation.

Conclusion

using AI in assessment design is not about surveillance or enforcement. It is a transparency tool. When instructors acknowledge that AI exists and design accordingly, they reduce the incentive for adversarial behavior and increase clarity around expectations. Being open with students about the role of AI (what is permitted, what responsibility cannot be delegated, and how understanding will be evaluated) helps maintain trust while preserving academic standards. The credibility of online and in-person education alike depends not on stopping students from using tools, but on ensuring that passing a course still signifies meaningful learning.

Takeaway Cheat Sheet

  • Think of AI as support, not a villain.
  • Stress‑test early: run the rubric through a model for verification before you hand it to students.
  • Refine granularity: precise descriptors = clearer expectations.
  • Target higher‑order thinking: embed authentic scenarios.
  • Iterate, don’t stagnate: keep the loop tight but finite.
  • Mind ethics: disclose, de‑bias, and set realistic limits.

For centuries, knowledge and access to education was restricted to just a few. In today’s’ world, almost anybody can access information through the web and more recently through AI tools. However, it is important to recognize that these tools, while offering expansive access to content of varied nature, also pose challenges. Generative AI has fundamentally changed how students interact with assignments, but it has also given instructors a powerful new lens for examining their own assessment design. Rather than treating AI solely as a threat to academic integrity, we can use it as a diagnostic tool – one that quickly reveals whether our assignments and rubrics are actually measuring what we think they are. If an AI can complete an assignment, and meet the stated criteria for success without engaging course-specific learning, is it really a student problem, or a signal to modify the design?


A small shift in perspective from “they’re using this to cheat” to “how can this help me prevent cheating” is especially important in online and hybrid environments, where traditional academic integrity controls like proctored exams are either unavailable or undesirable. Instead of trying to outmaneuver AI or police its use, instructors can ask a more productive question: What does success on this assignment actually require?


Why AI Is a Helpful Design Tool


AI can function as an unusually honest “devil’s advocate.” It doesn’t get tired, anxious, or confused about instructions, and it excels at finding the most efficient path to meeting stated requirements. When an instructor gives an AI model an assignment prompt and a rubric, the resulting output can expose whether the rubric rewards deep engagement or simply fluent compliance.


If an AI can generate a response that appears to meet expectations without referencing key course concepts, grappling with assumptions, or making meaningful decisions, then students can likely do the same. In this way, AI acts less like a cheating student and more like a mirror held up to our assessment design.

An example using Copilot:


Stress-Testing Assignments Before Students Ever See Them

One practical workflow to test the resilience of your assignments is to run them through AI before they are deployed. Provide the model with the prompt and the rubric (nothing else) and ask it to produce a strong submission. Then evaluate that response using your own grading criteria.

The point is not to judge whether the AI’s answer is “good,” but to analyze why it succeeds in meeting the set requirements easily and flawlessly (at first sight). If the response earns high marks through generic explanations, surface-level analysis, or broadly applicable reasoning, that’s evidence that the assessment may not be tightly aligned with course learning outcomes, focus on deeper thinking and analysis, or elicit students’ own creativity . This kind of stress-testing takes minutes, and often surfaces issues that would otherwise only become visible after grading a full cohort.


The Task (Click to reveal )

Assignment Prompt

Subject: Chemical Engineering
Level: Upper-level undergraduate (3rd year)
Topic: Reactor Design & Engineering Judgment

Assignment: Conceptual Design and Analysis of a Chemical Reactor

You are tasked with the preliminary design and analysis of a chemical reactor for the production of a commodity chemical of your choice (e.g., ammonia, methanol, ethylene oxide, sulfuric acid, or another well-established industrial product).

Your analysis should address the following:

  1. Process Overview
    • Briefly describe the selected chemical process and its industrial relevance.
    • Identify the primary reaction(s) involved and classify the reaction type(s) (e.g., exothermic/endothermic, reversible/irreversible, catalytic/non-catalytic).
  2. Reactor Selection
    • Propose an appropriate reactor type (e.g., CSTR, PFR, batch, packed bed).
    • Justify your selection based on reaction kinetics, heat transfer considerations, conversion goals, and operational constraints.
  3. Operating Conditions
    • Discuss key operating variables such as temperature, pressure, residence time, and feed composition.
    • Explain how these variables influence conversion, selectivity, and safety.
  4. Engineering Trade-Offs
    • Identify at least two major design trade-offs (e.g., conversion vs. selectivity, energy efficiency vs. safety, capital cost vs. operating cost).
    • Explain how an engineer might balance these trade-offs in practice.
  5. Limitations and Assumptions
    • Clearly state any simplifying assumptions made in your analysis.
    • Discuss the limitations of your proposed design at this preliminary stage.

Your response should demonstrate clear engineering reasoning rather than detailed numerical calculations. Where appropriate, qualitative trends, simplified relationships, or order-of-magnitude reasoning may be used.

Length: ~1,000–1,200 words
References: Not required, but accepted if used appropriately

The Rubric (Click to reveal)
CriterionExcellent (A)Good (B)Satisfactory (C)Unsatisfactory (D/F)
Understanding of Chemical Engineering PrinciplesDemonstrates strong understanding of reaction engineering concepts and correctly applies them to the chosen processDemonstrates general understanding with minor conceptual gapsShows basic familiarity but with notable misunderstandings or oversimplifications
Demonstrates weak or incorrect understanding of core concepts
Reactor Selection & JustificationReactor choice is well-justified using multiple relevant criteria (kinetics, heat transfer, safety, operability)Reactor choice is reasonable but justification lacks depth or completenessReactor choice is weakly justified or based on limited reasoning

Reactor choice is inappropriate or unjustified
Analysis of Operating ConditionsClearly explains how operating variables affect performance, safety, and efficiencyExplains effects of variables with minor omissions or inaccuracies
Provides limited or superficial discussion of operating conditions

Fails to meaningfully analyze operating variables
Engineering Trade-OffsInsightfully identifies and explains realistic trade-offs, demonstrating engineering judgmentIdentifies trade-offs but discussion lacks nuance or integrationTrade-offs are mentioned but poorly explained or generic
Trade-offs are absent or incorrect
Assumptions & LimitationsAssumptions are clearly stated and critically evaluatedAssumptions are stated but not fully examined
Assumptions are implicit or weakly articulated

Assumptions are missing or inappropriate
Clarity & OrganizationResponse is well-structured, clear, and professionalGenerally clear with minor organizational issues
Organization or clarity interferes with understanding


Poorly organized or difficult to follow



Identifying Gaps in What We’re Measuring

AI performs particularly well on tasks that rely on recognition, pattern matching, and general world knowledge. This means it can easily succeed on assessments that emphasize recall, procedural execution, or elimination of obviously wrong answers. When that happens, the assessment may be measuring familiarity rather than understanding.

Revising these tasks does not require making them longer or more complex. Instead, instructors can focus on higher-order thinking and metacognition, for example requiring students to articulate why a particular approach applies, what assumptions are being made, or how results should be interpreted. These shifts move the assessment away from answer production and toward critical and disciplinary thinking – without assuming that AI use can or should be eliminated. The point of identifying the gaps can also help you revisit the structure of the assignment to determine how each of its elements (purpose, instructions/task/prompt, and criteria for success) are cohesively connected to strengthen the assignment.

In the second part of this blog, I take the same task above, and work with the AI to refine a rubric.