Can You Co-Author This Paper?

// SLIDE 01 — HOOK

This paper is about your degree. And what it might be missing.

65–75%

of required master's coursework may map to Tier 1 — the tier where AI is now superhuman

The paper audits the top 10 STEM master's programs in the U.S. — the ones most OPT students attend. It asks whether those programs teach what H-1B-sponsoring employers actually require. You are both the subject of this research and a potential author of it.

// SLIDE 02 — LEARNING OUTCOMES

What you'll be able to do when this is done.

By the end of this, you will be able to…	Bloom's Level
Summarize the paper's argument and its stakes for OPT students specifically	Understand
Classify the paper's open TODOs by type and match them to your existing skills	Analyze
Apply CRITIQ's /brainstorm to generate at least one substantive contribution idea	Apply
Evaluate your co-authorship fit against the CRediT taxonomy	Evaluate
Decide whether to pursue co-authorship — and if yes, name the specific TODO you will own	Create / Decide

// SLIDE 03 — THE CORE CLAIM

The paper makes one claim. Here it is.

Programs teach what they've always taught. Employers now require something else. This paper documents the distance between those two facts. // working hypothesis — treated as hypothesis until the curriculum audit produces data

01

What programs teach

Tier 1 — pattern execution, algorithm implementation, framework application. Observable in syllabi through verbs: implement, build, apply, use, write, construct.

02

What H-1B sponsors require

Tier 3–5 — system design judgment, AI auditing, stakeholder navigation, causal reasoning. Appears primarily in electives, if at all.

03

What the paper does

Documents that mismatch with enough methodological rigor to publish — and recruits co-authors to close the open TODOs.

// SLIDE 04 — THE TIER FRAMEWORK

Five tiers. You need to know three of them.

T1

Tier 1 — AI is superhuman

Pattern execution, algorithm implementation, framework application. Observable in syllabi: implement, build, apply, use, write, construct. This is where most required coursework lives.

T3

Tier 3 — AI simulates, does not possess

Communication, stakeholder navigation, ethical judgment, cross-functional collaboration. Observable through: communicate, negotiate, advise, present, advocate.

T4

Tier 4 — AI is poor

Problem formulation, AI output auditing, system design under uncertainty, interpretive judgment. Observable through: evaluate, audit, formulate, critique, distinguish. Employers require this. Curricula rarely do.

T5

Tier 5 — Nearly absent from required coursework

Causal and counterfactual reasoning — root cause analysis, interventional thinking. The paper codes every required course at every audited institution against this framework.

// PAUSE & REFLECT — 01 OF 02

Stop. Think. Your courses, not this paper's.

PAUSE.

Think about your own required coursework. Which of your core courses would you code as Tier 1? Which — if any — would you code as Tier 3 or 4? What verbs appeared in the course descriptions?

// CHECK FOR UNDERSTANDING — 01 OF 02

// SLIDE 06 — OPEN TODOs

The paper has open TODOs. They're not all the same kind of work.

Literature & Source Work

Read and verify primary sources — CS2023 guidelines, SEVIS data, Anderson & Krathwohl (2001), prior Bloom's audit studies. Requires library access and careful reading of primary sources, not summaries.

Data Collection

Pull degree requirements, course descriptions, and learning outcomes from ten institutional websites. Archive with timestamps. Structured, replicable, and currently the bottleneck.

Coding — Second Coder Role

Apply the Bloom's-to-tier rubric to required courses. A second coder is a named author role — inter-rater reliability requires two independent coders. No prior research experience required.

Statistical Analysis

Chi-square tests, Cohen's Kappa, Cramér's V. Requires comfort with R or Python's sklearn — or willingness to learn it with documentation and a real dataset.

Writing

Methods, Results, Discussion — written after the analysis is locked, in a specific sequence. Requires academic writing experience, or willingness to develop it using CRITIQ on a real paper.

Most students assume co-authorship means writing. The paper needs coders and data collectors first. That's where it stalls — and where you can unblock it.

// SLIDE 07 — CRediT TAXONOMY

Co-authorship is not one thing. CRediT names fourteen.

CRediT Role	What it maps to in this paper
Formal Analysis	Coding the Bloom's-to-tier rubric; running chi-square, Cohen's Kappa, Cramér's V on the full dataset
Data Curation	Pulling and archiving degree requirements, course catalogs, and learning outcomes at all ten institutions with timestamps
Investigation	Verifying primary sources — CS2023 guidelines, SEVIS data, Anderson & Krathwohl, prior Bloom's audit studies
Methodology	Designing or refining the Bloom's-to-tier coding rubric, anchor examples, and adjudication notes
Writing – Original Draft	Methods, Results, Discussion — written in sequence after data is locked and coded
Writing – Review & Editing	Reviewing and tightening any section; running CRITIQ on drafted content before submission

One role, done rigorously, is a legitimate co-authorship claim. The orange rows are the current bottlenecks — the paper cannot move forward without them.

// SLIDE 08 — SELF-ASSESSMENT

Qualification is not a feeling. It's a match between skills and TODOs.

Data Collection You are qualified if you can navigate a university course catalog, save a webpage, and record a URL with a date stamp. // confidence: high — most master's students qualify

Coding — Second Coder Role You are qualified if you can read a rubric, apply it consistently, and discuss disagreements honestly. No prior research experience required — only precision and availability. // confidence: high — this slot is open right now

Statistical Analysis You are qualified if you've run a chi-square test or Cohen's Kappa in R or Python, or are willing to learn it from documentation with a real dataset in hand. // confidence: self-assess against your methods coursework

Literature Verification You are qualified if you have institutional library access and can read a primary source without relying on someone else's summary or a synthesis tool. // confidence: medium — requires access + disciplined reading

Writing You are qualified if you can produce a first draft that survives peer review — or are willing to use CRITIQ to build that skill on this paper, from real data. // confidence: self-assess — the next two slides are the test

// SLIDE 09 — THE CRITIQ HANDOFF

Paste the outline into CRITIQ. Run /brainstorm. See what you find.

01

Open CRITIQ

CRITIQ is a peer review and paper development tool. It reviews manuscripts and builds them from raw ideas using the same standard it would apply as a reviewer. It will teach you as it works.

02

Paste the full paper outline

The full outline — including TODO annotations, research questions, methodology section, and the working hypothesis. All of it. CRITIQ needs the complete context to find the real gaps.

03

Type /brainstorm

CRITIQ will ask: what did you notice? What's your gut explanation? What would prove you wrong? It will move you from observation to testable research question in six questions.

04

Identify one contribution

Look for: one gap, one methodology question, one thing the outline doesn't address that should be. If you can't find anything after running /brainstorm — that's data too. This paper may not be the right fit.

// SLIDE 10 — CRITIQ OUTPUT

/brainstorm turns an observation into a research question. Here's the structure.

CRITIQ Output Section	What it gives you
Your Research Question	One clean, specific, answerable question — not a topic. This becomes your entry point for a co-authorship conversation.
Your Working Hypothesis	One sentence: what you think is true and why. Sharpened by CRITIQ — but the observation is yours.
The Gap You're Filling	What's known, what's not, and where your question sits between the existing literature and the paper's open TODOs.
Study Design (Rough)	The simplest viable approach to test your hypothesis — constrained to what's actually feasible given the paper's scope.
What to Read Next	Three search terms or topic areas that extend understanding of the gap — not fabricated citations.
Concepts to Learn	Terms from the brainstorm to understand before drafting. Each is a /learn shortcut in CRITIQ.

This output is yours — not the paper's. A working hypothesis may be wrong, under-specified, or already addressed in the outline. That's fine. It's the beginning of a conversation, not the end of one.

// PAUSE & REFLECT — 02 OF 02

Stop. Decide. Before the next slide.

DECIDE.

After running /brainstorm: what was the one observation or gap you found? Can you name the CRediT role it maps to? Are you willing to own that contribution through to completion — not just through the interesting parts?

// CHECK FOR UNDERSTANDING — 02 OF 02

// SLIDE 12 — THE DECISION

Two answers. Both are correct. Only one is yours.

Yes — I'm In

Email the lead author with your /brainstorm output, the specific TODO you're claiming, the CRediT role it maps to, and your timeline

The second coder slot is open right now. If that's the only thing you're ready to do — that's enough to reach out. Send the email.

One role, done rigorously, to completion. That's the co-authorship standard. Not enthusiasm — completion.

Not Yet / No

Name what's missing — skill, time, access, confidence. That's not a no. It's a scoped gap you can close before the next TODO opens up.

You've read a real research outline and evaluated your fit against it honestly. That skill transfers to every paper you read from here.

A no now is not a no forever. The paper has a publication timeline — reach out when the gap closes. The TODO list will still be there.

// SLIDE 13 — CLOSE

Your curriculum has a gap. So does this paper. Pick one to close.

CO-AUTHOR IT.

// OWN A TODO // RUN /brainstorm // SEND THE EMAIL

The paper documents a structural mismatch between what STEM master's programs teach and what H-1B-sponsoring employers require. You are both its subject and its potential author. The /brainstorm command is the entry point. The second coder slot is open. The TODO list is documented.

What's missing is a decision.

Irreducibly Human // Nik Bear Brown // Research Methods // 2025