Screen 1 of 90:00–0:05
Six claims. One rule.
Four roles, one path. Round 4 is the mentor, who asks what your evidence establishes.
1AUTHOR
2VERIFIER
3READER
4MENTOR
Relay order: AUTHOR, then VERIFIER, then READER, then MENTOR.
“Say only what the evidence says.”
Every case is course-owned and synthetic.
One laptop per table. Choose, read the consequence, pass the claim on, then write the rule.
Equal-credit starting points: no prior packet, public link, live web, AI account or external lesson is required. Course cards, projected or transcript routes, paper, LMS text and approved accessible routes earn full credit.
Why this leads forwardCase 1: the answer came back in two seconds.
5 MIN
Screen 2 of 90:05–0:12
CASE 1 - THE ROUTE - speed is not evidence.
The answer came back in two seconds. Where did it come from?
Seven minutes, four rounds, then your rule.
1AUTHOR
2VERIFIER
3READER
4MENTOR
Relay order: AUTHOR, then VERIFIER, then READER, then MENTOR.
The case. A club member asks whether the Study Sprint prototype helps people focus. An assistant answers in two seconds, citing a 2025 study. Nobody has opened the app or the study. The answer is due at Thursday's meeting.
Model clause. A claim may use any appropriate route, must carry the route it came from and the limit that remains, and may never treat a fast answer as evidence.
Where the roles disagreed. The author wanted an answer by Thursday, the verifier a checkable one, the reader something to act on, the mentor something defensible.
No live web, AI account or prior packet is required. Course cards, projected, paper and accessible routes earn equal credit.
Why this leads forwardNext case: the page that vouches for itself.
7 MIN
Screen 3 of 90:12–0:19
CASE 2 - THE PAGE - a page can only tell you what it says about itself.
Everything on the page checks out. What did you not check?
Seven minutes, four rounds, then your rule.
1AUTHOR
2VERIFIER
3READER
4MENTOR
Relay order: AUTHOR, then VERIFIER, then READER, then MENTOR.
The case. A site says the prototype cut study time by a third. The page has a methods section, a chart, a named author and a 2025 date. Everything on the page supports everything else on the page.
Model clause. A claim may start on the page that makes it, must carry one independent source checked away from that page, and may never rest on a page vouching for itself.
Where the roles disagreed. The author wanted a usable number, the verifier an outside check, the reader something repeatable, the mentor a sentence that survives scrutiny.
Course-supplied source cards are the required route. No prior packet, public link, live browsing or external lesson is needed for full credit.
Why this leads forwardNext case: the citation that isn't.
7 MIN
Screen 4 of 90:19–0:26
CASE 3 - THE CITATION - four questions hide inside one citation.
The citation is perfectly formatted. Does the source exist, and does it say this?
Seven minutes, four rounds, then your rule.
1AUTHOR
2VERIFIER
3READER
4MENTOR
Relay order: AUTHOR, then VERIFIER, then READER, then MENTOR.
The case. An assistant produced a sentence with a citation: Hartley and Nwosu, 2025. The formatting is perfect. No record of that paper exists in any catalogue the class can reach, and the sentence it supports claims a grade improvement.
Model clause. A claim may cite and disclose separately, must carry a findable source that supports this exact sentence, and may never use citation-shaped text as proof.
Where the roles disagreed. The author wanted the point kept, the verifier a source that exists and fits, the reader two clean records, the mentor traceability.
A source credible for one question can be irrelevant to another. Never submit an unchecked citation, a private chat or account screen.
Why this leads forwardNext case: the status.
7 MIN
Screen 5 of 90:26–0:33
CASE 4 - THE STATUS - blocked is a check condition, not a label.
You could not open the source. Is that a fifth status?
Seven minutes, four rounds, then your rule.
1AUTHOR
2VERIFIER
3READER
4MENTOR
Relay order: AUTHOR, then VERIFIER, then READER, then MENTOR.
The case. Your sentence says the prototype produces a two, twenty, three plan for a given input. You ran it once and saw that. The source describing the wider claim sits behind a login the class cannot reach.
Model clause. A claim may be observed, source-supported, inferred or unsupported, must carry the status its evidence warrants, and may never be upgraded because a check was blocked.
Where the roles disagreed. The author wanted the strongest honest label, the verifier a rule for blocked checks, the reader something to trust, the mentor calibration.
Status applies to the exact sentence, not to the student, the project, the source, or the quality of the writing.
Why this leads forwardNext case: the trail.
7 MIN
Screen 6 of 90:33–0:40
CASE 5 - THE TRAIL - a bibliography is not a mapping.
You listed six sources. Which sentence does each one support?
Seven minutes, four rounds, then your rule.
1AUTHOR
2VERIFIER
3READER
4MENTOR
Relay order: AUTHOR, then VERIFIER, then READER, then MENTOR.
The case. The packet lists six sources in a bibliography and makes nine claims in the text. A reader asks which source supports the sentence about cost. Nobody can answer it without rereading all six.
Model clause. A claim may draw on several sources, must carry one row mapping this sentence to what supports it, and may never offer a bibliography in place of that mapping.
Where the roles disagreed. The author wanted the packet finished, the verifier one row that holds, the reader a sentence to follow, the mentor auditability.
A privacy-safe evidence unit is a short transcript, a class-safe excerpt or an observed check. Never attach the private archive.
Why this leads forwardNext case: what it does not prove.
7 MIN
Screen 7 of 90:40–0:47
BACK TO case 1: the route this story rests onCASE 6 - WHAT IT DOES NOT PROVE - true sentences can build a false story.
Every sentence is true. Is the story they add up to true?
Seven minutes, four rounds, then your rule.
1AUTHOR
2VERIFIER
3READER
4MENTOR
Relay order: AUTHOR, then VERIFIER, then READER, then MENTOR.
The case. Every sentence in the story is true. Together they suggest a tested product that improved learning. A mentor asks what the evidence actually establishes.
Model clause. A project story may describe what was built, must carry what the evidence does not establish, and may never imply mastery, measured impact or readiness that no evidence supports.
Where the roles disagreed. The author wanted to sound capable, the verifier a true whole, the reader a clear picture, the mentor a survivable story.
Never submit an invented verification, an unchecked citation, a private chat or archive, a resume, an application, or identity evidence.
Why this leads forwardNext: turn the six rules into one auditable source trail.
7 MIN
Screen 8 of 90:47–0:59
Build a trail another reader can audit.
Use one row per exact claim. A source list is not a mapping.
Build one complete row plus disclosure. The source record answers where material came from; the disclosure answers how help shaped the work. Neither replaces the other. A blocked check names the barrier and the next check; it does not upgrade the status.
Privacy-safe evidence unit: a short transcript, class-safe excerpt, or instructor-observed check. Never attach the private archive.
Why this leads forwardThe close carries this row, and your six rules, into Meeting 2.
12 MIN
Screen 9 of 90:59–1:08
Leave with one claim you can defend.
Your table wrote six rules today. Meeting 2 turns this evidence into a career story that has to satisfy them.
One sentence tied to the actual evidence route.
Observed / sourced / inference / unsupported.
Source + additional check or blocked reason.
What remains untested.
Never submit: an unchecked citation, an invented verification statement, a full private chat, a private draft or archive, a paywalled or restricted upload, a resume, a job application, contact details, grades, account screens, or identity evidence.
Equal credit: the supplied cards, a projected or transcript route, paper, LMS text and approved accessible routes all earn full credit. No prior packet, public link, live web or AI account is required.
Why this leads forwardMeeting 2 maps the verified work — not adjectives — to career evidence.
9 MIN
Screen 1 of 110:00–0:05
What could you show an employer from this course?
Bring the exact claim, status, trail, disclosure, and limitation from Meeting 1. If unavailable, recreate them with the Study Sprint case.
Which field most directly stops a polished story from becoming an overclaim? Defend your answer, then identify the field that makes the story auditable.
Why this leads forwardThe task-bundle lens identifies where AI helped and where human responsibility remained.
5 MIN
Screen 2 of 110:05–0:11
Break the project into tasks. Who did each one?
Classify each task by appropriate assistance and decision authority. Tool capability does not transfer responsibility.
Classify your own workflow: one assisted task, one human decision, and one rejected or out-of-scope task. Explain why the decision authority stays human.
Why this leads forwardCareer competencies describe demonstrated behaviors, not tool ownership.
6 MIN
Screen 3 of 110:11–0:18
Choose one competency the evidence can support.
NACE, the National Association of Colleges and Employers, publishes eight career-readiness competencies. This activity uses four relevant lenses and does not claim mastery or use the paid assessment tool.
PROJECT
EVIDENCE
EVIDENCE
Select one lens: name the action, artifact, judgment, and limitation that demonstrate one behavior. Do not write “I mastered critical thinking.”
Adapted from the official NACE Career Readiness Competencies. NACE lists eight; this course selects four.
Why this leads forwardRound 1 removes unsupported adjectives from a project story.
7 MIN
Screen 4 of 110:18–0:25
Round 1 of 3: Replace an adjective with evidence.
Weak story: “I created an innovative AI app that dramatically improved student productivity.”
Which revision is most defensible?
Choose one answer; that click commits your judgment and reveals its evidence rationale. Then revise the packet.
Record: action + artifact + exact evidence; remove innovation, impact, and mastery claims unless independently supported.
Why this leads forwardRound 2 maps the evidence to one competency behavior.
7 MIN
Screen 5 of 110:25–0:32
Round 2 of 3: Build the evidence map.
A career claim needs the same traceability as a research claim.
Build your map: exact input/action → observed or supplied output → route/status → limitation. A two-row transcript or instructor-observed check is enough; do not attach a private underlying file.
Why this leads forwardRound 3 turns the map into one self-contained paragraph.
7 MIN
Screen 6 of 110:32–0:39
Round 3 of 3: Write a credible project story.
Include purpose, action, evidence, human judgment, assistance disclosure, and limitation. Do not include personal contact or application details.
Write 90–130 words. Which sentence carries the strongest evidence, and which sentence prevents overclaiming?
1 · COMMITWrite one claim or decision and the evidence behind it.
2 · CHALLENGEAsk what the evidence does not establish.
3 · REVISENarrow the claim or add the missing check.
Working alone? Write the strongest support and the strongest limitation, then decide.
Course clarification
A strong story names the bounded prototype, an inspectable action/test/source trail, one human review decision, any material AI assistance, and a limitation. It may connect that behavior to one competency, but it does not claim mastery, measured impact, employment readiness, or production deployment.
Why this leads forwardA peer, solo, projected, or written explanation now tests whether the story can be defended.
7 MIN
Screen 7 of 110:39–0:46
Compare two stories by evidence—not confidence.
Story A is polished but claims impact. Story B is modest and traces one test, source decision, disclosure, and limit.
“I built an innovative AI app that dramatically improved student productivity.”
No test, source, judgment, disclosure, or limitation.
“I configured and tested a timed-task prototype, verified its 2/20/3 output for one named run, rejected an invented grade claim, disclosed AI help with wording, and recorded that learning impact remains untested.”
Action, artifact, judgment, disclosure, and limitation are inspectable.
Which story better supports Communication, Critical Thinking, or Technology? What one revision would strengthen the weaker story without inventing new evidence?
1 · COMMITWrite one claim or decision and the evidence behind it.
2 · CHALLENGEAsk what the evidence does not establish.
3 · REVISENarrow the claim or add the missing check.
Working alone? Write the strongest support and the strongest limitation, then decide.
Course clarification
Story B is more defensible because readers can inspect its action, artifact, decision, disclosure, and limitation. Strengthen Story A by removing unmeasured impact and replacing it with a named artifact or observed result. Style cannot substitute for evidence.
Why this leads forwardExplanation practice tests whether students understand the evidence chain.
7 MIN
Screen 8 of 110:46–0:53
Explain the project without exposing personal information.
Choose any route: partner exchange, private written rehearsal, instructor-projected synthetic case, speech-to-text, or approved accessible equivalent.
QUESTION 1What exact problem and boundary did the project address?
QUESTION 2Which artifact or source supports your strongest claim?
QUESTION 3What did you personally verify or reject, and what remains untested?
Listener response: “The strongest evidence I heard was ___; one claim to narrow or clarify is ___.”
This is classroom explanation practice, not a job interview, public presentation, or personality assessment.
Why this leads forwardA privacy check removes evidence that is unnecessary for the learning target.
7 MIN
Screen 9 of 110:53–0:59
What must your evidence include, and what must it never include?
The assignment assesses claim verification and explanation. It does not need a private resume, job application, contact detail, account screen, or personal history.
POSSIBLE EVIDENCE
Run the filter: keep only what supports the learning target. Your own name is allowed only in the packet header; browser draft saving excludes it.
Why this leads forwardThe five-page packet preserves evidence without collecting the underlying private archive.
6 MIN
Screen 10 of 110:59–1:04
Assemble five pages of research and career evidence.
Most fields are drafted during class. The packet grades claim calibration, source traceability, disclosure, human judgment, and explanation—not tool or account access.
PAGE 1Claim inventory + disclosure.
PAGE 2Source trail.
PAGE 3Lateral reading + independent check.
PAGE 4Career map + 90–130 word story.
PAGE 5Explanation, privacy, limit, Topic 12.
Accepted routes: HTML, LMS text, DOCX/export, five-page PDF, scanned paper, speech-to-text, projected/transcript evidence, or approved accessible equivalent. No live web or AI use is required.
Why this leads forwardThe close carries one weak sentence into Topic 12 prompt-strategy practice.
5 MIN
Screen 11 of 111:04–1:08
Carry one evidence problem—not a polished answer—into Topic 12.
Topic 12 will improve prompt strategy. It needs a bounded task, context, output format, and evaluation rule from this week.
What needs revision?
What must the prompt preserve?
What must it not invent or expose?
How will a human check the revision?
Submit before Topic 12: use the LMS calendar for the exact date/time/timezone, late rule, and resubmission rule. Bring or recreate the four-field handoff regardless of packet status.
Why this leads forwardTopic 12 turns this evidence contract into stronger prompting and evaluation.
4 MIN