How You can Apply College Research Writing Skills to AEO
Once, in an attempt to make finding sources more tactile, I worked with an academic librarian to create a scavenger hunt.
Students broke into teams and had to:
- Use databases to find books and articles
- Use a book’s index to answer questions
- Use bibliographies to track down additional sources
They got competitive, and it was fun to watch them running through the stacks, opening books that probably hadn’t been touched in decades.
That was the easy part.
Getting them to evaluate whether those sources were actually credible? That was the real challenge.
In my last blog post, I mentioned a student using a third-grade teacher’s classroom blog, complete with student observations, as a source on a literary term.
He genuinely couldn’t understand why that didn’t qualify. And to be fair, it was relevant. It answered the question and was easily discoverable – I would still love to know how Ms. Smith’s blog ranked so highly.
That’s what had him convinced it was a valid source. Teaching students that not all sources are equal is a constant battle. Add in database training, and I was lucky if they weren’t defaulting back to Wikipedia.
What we needed was a framework.
How I Taught Students to Evaluate Sources
The CRAAP Test helped students learn to separate convincing sources from credible ones.
Currency: Is it current? When was it published/updated?
Context matters. For students researching the treatment of women’s mental illness in Charlotte Perkins Gilman’s “The Yellow Wallpaper,” a source from the 1950s isn’t just outdated; it could be entirely misleading.
Relevance: Does it actually answer the question?
Once upon a time, students loved Cliff’s Notes because they answered their questions quickly. But that was the problem. Cliff’s Notes told you what happened. A research paper needed to explain what it meant.
Authority: Who wrote it? What are their credentials? Who published it?
Ms. Smith’s third-grade blog is a simple but real example. Wikipedia and its anonymous, everyman authorship was another constant battle. I always asked the same question: Why should we trust this person/source?
Accuracy: Can you verify the information? Are sources cited?
If a claim didn’t trace back to a real source, students learned to see it as a red flag. Credible sources show their work.
Purpose: Why does this exist? Is it trying to inform, sell, persuade?
A .com selling supplements will tell you different things about vitamin D than a peer-reviewed medical journal. Both might include accurate information, but their purposes shape what they emphasize and what they leave out
I taught that framework for 17 years. Lately I’ve been realizing we’re using it again, just in a different context.
E-E-A-T (What LLMs Evaluate)
This is how LLMs decide what’s credible enough to include in an answer.
Experience: Does this reflect real, lived insight?
That’s why platforms like Reddit and forums show up so often. They’re messy, but they’re human.
In an AI-driven world, that human element becomes more valuable, not less.
Expertise: Does the source actually know what it’s talking about?
We still want authority; we just don’t trust it blindly.
A university blog post and a student’s TikTok can both be useful. They answer different kinds of questions.
Authoritativeness: Is this a recognized source?
A .edu domain helps, but it isn’t enough. Are you actually cited? Referenced? Part of an ongoing conversation?
Trustworthiness: Is this consistent and transparent?
If your site says one thing and the rest of the internet says another, that gap matters. LLMs notice the disconnect and report on it.
LLMs are essentially running the CRAAP test at scale but with some key differences.
Where Academic and AI Credibility Diverge
1. Authority is no longer top-down
In academia, credentials carry the most weight.
Now, authority is layered:
- Institutional credibility
- Expert knowledge
- Community validation
A published article and ten Reddit threads can both be credible depending on the question.
2. Citations Matter but Differently
It’s still about being part of the conversation.
But not all citations carry equal weight. An endless loop of press releases doesn’t signal authority.
Being referenced by credible, independent sources does.
3. Recency Matters More
Academic work can stay relevant for decades. Search doesn’t work that way.
If content doesn’t reflect the current landscape, it loses value, especially in fast-moving fields.
4. Transparency is Everything
“Cite your sources” isn’t new. But now it’s visible and testable at a larger scale than ever.
If your claims don’t hold up across the ecosystem, it shows.
What This Means for Higher Ed
Universities already have everything LLMs are looking for.
- Built-in authority
- Actual subject matter experts
- Original research
- Established credibility
Yet most higher ed content doesn’t reflect that.
Instead, it leans on:
- Generic marketing language
- Thin content
- Hidden expertise
- B2C tactics that flatten differentiation
The opportunity isn’t to create something new. It’s to use what already exists more intentionally.
Practical Application: The Source Credibility Audit
Run your content through both lenses:
The Student Test (CRAAP):
- Would you let a student cite this page in a research paper?
- If not, why? Fix those issues.
The LLM Test (E-E-A-T):
- Is there clear authorship and expertise?
- Are claims supported by data or sources?
- Is it actually useful or just descriptive?
- Would anyone credible reference this?
Quick wins:
- Add author bios with credentials to key pages
- Link to original research/data
- Build deeper, more complete resources
- Showcase faculty expertise on relevant topics
- Refresh outdated content
The Credibility Question
I used to ask my students one thing: Are you willing to stake your grade on this source?
That question hasn’t changed. What has changed is the scale.
If your content wouldn’t pass the test in a composition classroom, it probably won’t get cited by AI either.
The difference? LLMs are asking that question millions of times a day.