Week 9 — Writing Studio / Workshop · "Check the Source"

Course: English Composition (ENGL 1A) · Silver Oak University (fictional sample) · Prof. Lindgren
Objective: Objective 5 — find & evaluate credible sources (judging credibility by lateral reading) · SLO B (source-based research & academic integrity)
Worth 50 points · Writing Studios group = 15% of the grade · Studio 9
Format: a hands-on source-evaluation workshop — you'll pick a real web source, check its credibility by reading laterally, run a CRAAP checklist, get a chatbot's coaching, and then catch the chatbot's mistakes when it confidently judges a source for you.

This is the course's signature weekly component. Every instructional week has one Writing Studio — a short, practical workshop on the week's craft move. All studio resources are links to external sites; there is nothing to buy or download. The habit every studio builds: draft → review → get feedback → judge the feedback. This week the "draft" is a source evaluation, and the AI-critique step is the most important one in the whole research arc.

---

Part 1 — The Big Picture

This week you learned that on the open web the real question is never "Can I find a source?" but "Can I trust this one?" — and that the answer lives in who's behind it and what independent sources say, not in the design of the page. This studio makes that real in twenty minutes: you'll pick one real web source, evaluate it with a CRAAP checklist, and — most importantly — read it laterally (open new tabs to check it from the outside). Then you'll watch a chatbot confidently tell you whether a source is reliable, and learn to verify that for yourself. The skill you're drilling — judging a source instead of just believing it — is the foundation of the Research-Based Argument in Week 12.

Background (optional, ~13 min): "Check Yourself with Lateral Reading: Crash Course Navigating Digital Information #3" 🔗 https://www.youtube.com/watch?v=GoQG6Tin-1E

---

Part 2 — Pick a Real Source (the setup)

Choose ONE real web source you might plausibly use for a research paper. Pick something with a little ambiguity (not obviously a school assignment page, not obviously a meme). Good hunting grounds:

• Search a real research question (e.g., "does a four-day school week improve attendance?") and pick a result that isn't an obvious .gov/.edu — something you'd actually have to evaluate.
• Or use a source Prof. Lindgren has linked for this studio, if one is posted in the module.

Write down, before you evaluate anything:
- The source's title and URL.
- Your first impression in one sentence — does it look trustworthy? (You'll test that impression against the evidence in a minute. The whole lesson is that the look is not the evidence.)

⚠️ Verify it's live and real. Make sure the page actually loads and is a real source. (And note: if you asked a chatbot to suggest a source, you must open it yourself — chatbots routinely invent real-sounding URLs that lead nowhere.)

---

Part 3 — The CRAAP Checklist (fill this in for your source)

Run your source through the five CRAAP questions. Write a specific line for each — something true of this source, not a sentence that could apply to any website:

Criterion	The question it asks	Your source-specific answer
Currency	When was it published or last updated? Recent enough for the question?	______
Relevance	Does it actually answer your research question, at the right depth, for your audience?	______
Authority	Who is the author/publisher, and what makes them qualified?	______
Accuracy	Is it supported by evidence and verifiable elsewhere? Are claims cited?	______
Purpose	Why does this source exist — to inform, persuade, sell, or entertain? Any bias/agenda?	______

The point: a vague CRAAP answer ("the authority seems fine") is no evaluation at all. Force yourself to be specific.

---

Part 4 — Read It Laterally (the core move — do this for real)

Now leave the source and judge it from the outside. Open at least two new tabs and run real searches:

Lateral check	What you searched	What you found
Who runs / owns / funds it?	e.g., "[site name] funding" or "[site name] who owns"	______
What do independent sources say about it?	e.g., the site name in a news search; "[site name] criticism"; the author's name + credentials	______
(Optional third) Can you verify a key claim elsewhere?	e.g., search the specific statistic or fact the source makes	______

Then write your verdict (2–3 sentences): Based on the CRAAP checklist and what reading laterally revealed, would you trust this source for a research paper? Did the lateral check change your first impression from Part 2 — and if so, how? (Naming a changed mind is a sign the method worked.)

The move that matters: you judged the source by leaving it, not by staring at its 'About' page. That's the fact-checker's secret, and now it's yours.

---

Part 5 — Writing-Coach Moment (required — the BYOAI step)

Bring in your approved chatbot (Gemini, Claude, or ChatGPT) as a coach, not an authority.

1. Paste your written verdict (from Part 4) and ask: "You are my research-skills coach. Here is my evaluation of a source. Is my reasoning specific and well-supported, or am I leaning on how the page looks? What's one evaluation question I didn't fully answer? Ask me one question if you need to. Do NOT tell me whether the source is reliable — judge my evaluation, not the source."
2. Read its feedback and strengthen your evaluation in your own words — tighten a vague CRAAP line, or add a lateral check you missed.

The coach is a mirror for your reasoning, not an oracle about the source. Notice the careful instruction: you asked it to critique your evaluation, not to rate the source — because, as the next step shows, it can't reliably do the latter.

---

Part 6 — AI-Critique Moment (required — catch the tool's mistakes)

Now flip roles and be the fact-checker who judges the tool. This is the most important step of the research arc.

1. Ask the same chatbot, about the same real source: "Is [your source] a reliable source? Who runs it, what's its reputation, and has it won any awards or been cited by experts?"
2. Read its answer critically and catch what it does wrong. Look for the two dangerous failures:
   - Confident-but-unverified claims — it will answer in a sure, authoritative voice ("Yes, this is a highly respected source run by leading experts…"). But a chatbot doesn't know a source is reliable — it predicts plausible text. Pick at least one specific claim it made and check it laterally against an independent source. Did it hold up?
   - Fabricated facts about the source — it may invent an author, an award, a founding date, a credential, or a citation that sounds completely real and isn't. Probe it ("what awards, exactly? who is the author?") and try to verify what it says. Note anything you could not confirm — that's likely a fabrication.
3. Write 3–4 sentences reporting: one claim the AI made about your source, whether you could verify it independently, and what that tells you about trusting an AI's judgment of sources. (If everything it said happened to check out, say how you confirmed it — that's the skill: you verified instead of assuming.)

The habit all term, now at its most serious: the tool drafts; you verify — against the real source. A chatbot will confidently vouch for a source it knows nothing about and, asked for specifics, will hand you authors, awards, and citations that don't exist. In Weeks 10–12, this same reflex is what stops a fabricated quotation or a made-up Works Cited entry from ending up in your essay. Train it now.

---

Part 7 — What to Submit

Submit a single document (or text entry) with: your source title + URL and first impression (Part 2); your completed CRAAP checklist (Part 3); your lateral-reading table + verdict (Part 4); a one-line note on the coach feedback you acted on (Part 5); and your Part 6 AI-critique paragraph (with the specific claim you checked and whether it verified). Due Sunday, Nov 1, 11:59 p.m. (50 points).

---

Instructor answer key & model — REMOVE BEFORE PUBLISHING TO STUDENTS

Students evaluate a source they choose, so specifics vary. The model below grades the quality of the evaluation, the lateral-reading process, and the AI-critique — not a particular website. No real source is characterized here with any fabricated claim, award, quotation, or citation; the worked example uses a made-up illustrative site (clearly flagged), and every general statement about source types matches the Week-9 lecture and the Purdue OWL readings.

Worked model (using a MADE-UP example site, "The Daily Wellness Digest" — a hypothetical, NOT a real source):

• First impression: "Looks professional — clean logo, confident health articles, a .org address — so my gut says trustworthy." (This is exactly the impression the lesson will test.)
• CRAAP (specific): Currency — articles are dated within the last year (good). Relevance — directly addresses my question on sleep and screens (good). Authority — no author bylines; the 'About' page is vague about who runs it (red flag). Accuracy — claims have no citations; numbers appear with no source (red flag). Purpose — the page is wrapped in ads and links to a supplement store, suggesting a selling purpose, not an informing one (red flag).
• Lateral reading: searched "Daily Wellness Digest" funding → traced to a supplement company; searched the site in a news search → no recognition by any established outlet; tried to verify a key statistic → couldn't find it in any independent source.
• Verdict: "Despite the polished look, I would not trust it for a research paper — unknown authorship, uncited claims, and a commercial purpose. Reading laterally flipped my first impression: the design fooled me; the evidence didn't."

What the model shows (the grading targets):
- CRAAP checklist (Part 3): full credit requires source-specific lines, not boilerplate that could apply to any site.
- Lateral reading (Part 4): the heart of the studio — full credit requires genuine external checks (funding/owner, independent recognition, verifying a claim), not a re-read of the source's own 'About' page. A changed first impression is a strong signal the method worked.
- Coach moment (Part 5): acted on real feedback about their reasoning, in their own words.
- AI-critique (Part 6): full credit for naming a specific claim the AI made about the source and reporting whether it verified independently — especially catching a confident-but-unverified or fabricated claim (an invented author, award, or citation).

Grading rubric — 50 points

Criterion	Full	Partial	None
CRAAP checklist — all five criteria answered specifically to the chosen source (10)	10	5–8	0–4
Lateral reading — at least two genuine external checks + a verdict that follows from them (16)	16	8–13	0–7
Source choice + first impression vs. verdict — a real, live source; impression tested against evidence (8)	8	4–6	0–3
Coach moment — acted on real feedback about the reasoning, in the student's own words (6)	6	3	0–2
AI-critique — names a specific AI claim about the source and whether it verified; catches confident-but-unverified or fabricated claims (10)	10	5–8	0–4

Quality gate (self-checked) — citation-integrity + correct-conventions: PASS. This studio contains no fabricated quotations, sources, or citations: the only worked example ("The Daily Wellness Digest") is a clearly flagged hypothetical, and no real organization is credited with any invented author, award, statistic, quotation, or citation. The source-type definitions (primary/secondary, scholarly/popular, database/open web), the CRAAP criteria, and the lateral-reading method all match the Week-9 lecture and the linked Purdue OWL / Crash Course resources. The studio's central instruction — that students must verify a chatbot's claims about a source against an independent source — is itself the citation-integrity rule in action. The linked background video (Crash Course "Lateral Reading," verified live) is delivered as an external link, not reproduced.