Two best-practices to prevent difficulty with scientific authorship include: discussing authorship early and often, and using acknowledgment statements to clarify contributions. Newer trends like co-first authorship hold promise for more flexible structures in future.

The problem with authorship

Science is experiencing a reproducibility crisis that stems from systematic bias (1).

Systematic bias in medicine comes in many forms. It may mean selectively reporting positive results, doing selective analysis, or designing research with incorrect specifications (2). It can also come in writing, publishing, and authorship. The incentive structure of biomedical research has a pervasive influence on how and what biomedical research gets published (3).

High-profile scientists such as Marcia McNutt, former Editor-in-Chief of The Journal Science, have called for system-wide authorship reform (4).

In this article, we examine authorship bias and propose two simple practices to prevent both authorship bias and authorship disputes.

Many scientists are familiar with at least one well-documented instance of authorship fraud. However, authorship fraud is just an extreme example in the more subtle issue of authorship bias, which systematically corrupts the accuracy, equity, and quality of published clinical evidence (5,6).

Typical examples of authorship bias

In a 2011 survey, authors of nearly 900 articles at six high-profile medical journals — Annals of Internal Medicine, JAMA, Lancet, Nature Medicine, New England Journal of Medicine, and PLoS Medicine — estimated that 8% of their published studies lacked an author. A further 18% of studies had authors who had not met the criteria for authorship (7).

• Missing authors. Valid contributors intentionally or unintentionally omitted from scientific credit. This practice is a leftover from an era when authors of disadvantaged populations — minorities, the developing world and women— were unrecognized for their various scientific achievements (5, 8,9 ). One example of this form of selective credit was popularized in the 2016 movie Hidden Figures, which dramatizes female African-American mathematicians at NASA.

• Ghost authorship. The practice of omitting professional contributors from scientific credit. Ghost authorship is a growing issue in bioscience. It is related to an increase in professional medical writers and pharmaceutical-sponsored research.

• Honorary authorship. The practice of “guest,” “gift” or “legacy” authorship. This term refers to senior authors taking credit for a junior author’s work. It may also be an author included in a publication out of respect rather than legitimate scientific contribution. Credit confusion is one of the most common reasons for academic disputes, often involving a hierarchical power imbalance in medicine, academia, or other sciences.

Solutions for authorship

Scientific authorship has two simple functions: credit authors for their work, and document responsibility for its authenticity.

An accepted standard for authorship is the International Committee of Medical Journal Editor’s (ICMJE) 4 criteria:

1. “Substantial contributions to the conception or design of the work; or the acquisition, analysis, or interpretation of data for the work; AND

2. Drafting the work or revising it critically for important intellectual content; AND

3. Final approval of the version to be published; AND

4. Agreement to be accountable for all aspects of the work in ensuring that questions related to the accuracy or integrity of any part of the work are appropriately investigated and resolved.”

However, ICMJE criteria are often misapplied, used as evaluation criteria at the end of a research project, instead of when they are most valuable — at the beginning.

Proper scientific authorship combines credit with accountability. This means that good authorship is intrinsically bound to research execution. A first-author is accountable for all aspects of a research project, including ethics approval, data collection, data analysis, write-up, and submission of their work.

While authorship can be renegotiated as workflow and research groups change, it should never be unclear at any point in a research project for any member of the team.

Two simple research practices can help scientists clarify authorship and prevent embarrassing complications down the road:

1) Discuss authorship early and often

Be upfront, candid, and practical about authorship in your work group.

Researchers are more comfortable with their contribution to a project when authorship is coupled with research workflow, which in turn, helps prevents academic fraud.

Attributing proper authorship is an ongoing and responsibility. It is important when a research project is first being considered and can be used as an aid for defining research tasks. It should clearly spell out who does what as new work is assigned or evaluated.

Authorship is also relevant when clinical partnerships are formed, as research personnel come and go, or as publication dates change.

A good co-authorship agreement will cover all the essential points on authorship, but an attitude toward discussing authorship openly and without rancor as a prelude to contribution is more valuable than a written agreement to clinical research collaborations.

2) Put authorship on the front page of the drafted research work

Let contributors see their formal credit when writing starts.

The historical practice of allowing the PI to assign authorship when work is submitted is outdated.

With today’s cloud-based collaborative writing projects, coauthors have access to drafted work in real-time. By placing authorship on the first page of research drafts, coauthors have the opportunity to review and clarify their contributions to a research project right off the bat and apportion their time and energy appropriately thereafter.

The first page of every research draft should have a title and an authorship list as it will appear in the final published work. It should also have an ‘acknowledgments’ section to differentiate between authors and contributors. This way, all collaborators can see their work recognized and understand how it intersects with others.

For example:

Acknowledgements: Author A and Author B conceptualized the topic. Author A and Author C contributed to research design. Author D oversaw data collection performed by Author A. Author E performed data analysis. Author A led the writing of the article and was assisted by all authors. All authors reviewed, revised and approved of the final article.

The future of authorship

The practice of scientific authorship is quickly evolving. One way to accurately acknowledge the diversification of scientific contribution in the future is by using more advanced authorship metrics.

For now, we can start by applying the metrics we do have with greater finesse. Good authorship points to a future of good science.

References:

1. J. P. A. Ioannidis, Why most published research findings are false. , e124 (2005).

2. N. S. Young, J. P. A. Ioannidis, O. Al-Ubaydli, Why current publication practices may distort science. , e201 (2008).

3. M. K. McNutt, Transparency in authors’ contributions and responsibilities to promote integrity in scientific publication. Proc Natl Acad Sci USA. , 2557–2560 (2018).

4. Y. A. Shen, J. M. Webster, Y. Shoda, I. Fine, Persistent underrepresentation of women’s science in high-profile journals. (2018), doi:10.1101/275362.

5. S. Cummings, P. Hoebink, Representation of Academics from Developing Countries as Authors and Editorial Board Members in Scientific Journals: Does this Matter to the Field of Development Studies? Eur. J. Dev. Res. , 369–383 (2017)

6. D. M. Bennett, D. M. Taylor, Unethical practices in authorship of scientific papers. Emerg Med (Fremantle). 15, 263–270 (2003).

7. J. S. Wislar, A. Flanagin, P. B. Fontanarosa, C. D. Deangelis, Honorary and ghost authorship in high impact biomedical journals: a cross-sectional survey. BMJ. 343, d6128 (2011).

8. D. Sarna‐Wojcicki, M. Perret, M. V. Eitzel, L. Fortmann, Where Are the Missing Coauthors? Authorship Practices in Participatory Research. Rural Sociology. 82, 713–746 (2017).

9. S. Cummings, P. Hoebink, Representation of Academics from Developing Countries as Authors and Editorial Board Members in Scientific Journals: Does this Matter to the Field of Development Studies? Eur. J. Dev. Res. 29, 369–383 (2017).

10. R. Smith, Authorship: time for a paradigm shift? BMJ. 314, 992 (1997).

11. R. Smith, Authorship is dying: long live contributorship. BMJ. 315, 696 (1997).

This article was written at the request of Research Medics, a professional research writing website, and is reprinted with permission. Originally published at Researchmedics.com on October 22, 2021.

It’s a digital ninja playbook for those of us who don’t think might makes right. And yes, I do follow Ninja Tips for Healthy Living. Consider this article a hard-earned #11.

It’s written by me, for me, in solidarity for the times I’ve been specifically stolen from by corporations. Corporations that bullied me and exhausted my appeals and took things I needed that I’d worked really hard to earn.

This article is about defending ourselves as consumers against automated theft by corporations. And how to let the emotions of frustration and helplessness go afterwards.

The exact play

I have been on the receiving end of this exact play three times.

2002, Best Buy. I had saved for over a year to buy a $3,000 laptop. I worked hard for that money as a wildland firefighter (hotshot) in the LA foothills. When the laptop developed a hardware issue, a button was sticking, I sent it back to Best Buy for warranty repair. The box came back to me a few weeks later. I opened it. Inside was my precious computer in pieces, covered in sticky soda pop that had spilled across the keyboard. The laptop was dead. A technician in their service center had, obviously, been drinking a soda while working on it, knocked the can over onto the open machine, and — knowing exactly what they had done — closed it back into the shipping box and sent it to me anyway. No one at Best Buy would acknowledge it. The story was unprovable from where I was standing. A tech somewhere had done the math and concluded: who’s she going to tell, and who would believe her? They were right. I spent months in tears at their customer service desk. Then ate the $3,000.

2021, Airbnb. I ran my ass off during COVID, up every 2 hours for years, handling test results in two countries with 300 employees. When the vaccine came out, I scheduled a company retreat with our savings, which turned into a nightmare. We cancelled the $8,258 booking in Bali within the refund window because Indonesia’s borders had been closed to Americans the whole time. The host had lied to us in the booking: “There are no such regulations that Americans cannot enter Indonesia.” This was false and documented, but Airbnb simply took the money and refused to refund. The chargeback I filed was initially awarded, then reversed by my bank. Airbnb was deploying its legal team to dispute all refunds. By September, Airbnb told us we had to bring it to court to get the refund we were owed. Our company ate the $8,000, and my staff never got a well-earned vacation.

2026, Expedia. I booked a flight to Brazil with Expedia. They cancelled the flight and simply removed it from the interface. I had to buy a $1,000 emergency flight at the airport to get home so I didn’t miss an important business meeting. Eight chatbot agents over five weeks. Cases merged silently. Refund promised, then denied. Finally, in writing: “Please use Copa Airlines’ website to request a refund.” The merchant of record disclaiming refund duty to the supplier, then Copa ran the same play on my team.
Different decade, different platform, different continent. Same playbook, winning with brute mega-indifference against a weaker opponent.

But this time it’s an automated playbook.

The dark refund pattern

I call this the dark refund pattern. It’s a corporate play against a consumer, executed at scale. We’ve all experienced it. And in health insurance, it kills people, most recently Brian Thompson.

1. Gaslight on the underlying facts.

2. Refuse to apply the published policy.

3. Lose paperwork, merge cases silently, deny employee actions, break continuity.

4. Tell external regulators (BBB, DOT, AG) you’ll resolve it directly. Don’t.

5. Win chargeback representations using merchant-controlled documentation or specialized legal teams.

6. Close communication. Direct the customer to your registered agent for service of process.

What’s new in the last two decades isn’t the unwillingness to make customers whole. Companies have always preferred to keep your money. What’s new is the cost structure.

How dark refunds change in an AI era

When customer service was human, denying a claim cost the company something — a rep’s time, the cognitive load of saying no to a person who could become a bigger problem. Customers won by being annoying enough to be more expensive than the refund. The negotiation was real because both sides paid.

Automation, and the institutionalization of plausible deniability, changed the cost asymmetry. The platform’s cost of denial dropped to approximately zero. A chatbot loop costs nothing to run. A silent case merge takes a database write. A registered-agent redirect is a templated reply. “I can’t prove our tech did it” is free to say at corporate scale. Meanwhile, your cost of pursuit stayed human — your hours, your attention, your sanity, your willingness to relive the same story across eight escalation agents who can’t see what the previous seven said.

They are not breaking the rules. They are pricing the rules out of your reach.

How to beat AI-generated dark refunds

Here is what I have learned, having just run every move on the list inside a single weekend:

1. Get everything in writing immediately. Chat transcripts are designed to be merged, lost, and disclaimed. The moment a phone call or chat gives you anything useful, send an email summarizing it back to the company. Their email reply (or non-reply) becomes evidence. Screenshot every chat before closing the window.

2. Stop fighting the company. Start documenting for regulators. After two or three good-faith attempts, you have already lost the customer-service war. Continuing to argue with chatbots is feeding the asymmetry. The remaining value of your case is as evidence for a third party. Pivot accordingly.

3. File the chargeback under the right reason code. Not “services not rendered” — that loses to representment because the company can produce a policy document. File under “not as described” (Visa 13.3) or “misrepresentation” (Visa 13.5) if there was a material misstatement in the listing or the dispute process. This single tactical choice is often the difference between winning and losing.

4. Use the regulator that matches the company. Airlines and ticket consolidators → DOT (airconsumer.dot.gov). Any California-headquartered business → CA AG (oag.ca.gov/report). Anyone → BBB as a supplement, not a primary. State Attorney Generals complaints have no statute of limitations on filing, and they aggregate. Your single complaint contributes to a pattern the AG may already be tracking.

5. Frame the regulator filing as pattern documentation, not a personal grievance. “Please help me get my $944 back” is triaged as a low-priority individual complaint. “I am documenting a structured dispute-suppression pattern that may be relevant to your office’s ongoing oversight” is read. Lead with the structural claim, name the legal hooks (in California: Business & Professions Code §17200, §17500), and explicitly subordinate your personal loss.

6. Send a formal notice with a deadline before going public. A clear, professional email stating the facts, the reason codes, the regulators you have filed with, the consequential damages you are reserving, and a response window — usually 10 business days — does three things. It creates a documented final attempt at resolution. It triggers internal escalation processes that don’t exist in chat. And it gives you clean grounds to go public if they don’t respond.

7. Hold public posting as a ripcord, not a first move. A well-documented thread on X or a Substack post with screenshots is one of the few moves that still costs the company something at corporate scale. It also costs you something — attention, comment-section engagement, retaliation risk. Save it for after the deadline expires.

8. Recognize when to convert and stop. Not every loss can be recovered. The 2002 Best Buy file cannot. Some files are tuition payments toward learning the playbook. The point of doing the work above is not always to win — it is to make your file part of a record that eventually forces the math to change.

Once your file is filed, you are done. Let it go.

What this taught me

I consider this a modern analog of what Camus describes in The Stranger. Children with a rifle, kill a man on a beach — because people at a long distance look like dots.

And they aren’t.

They are people.

Globalization has taught us we don’t matter, and the truth is, we really DO matter. The single best business book I ever read was A Course in Miracles. To really get this right, and stop just trading victim-perpetrator-rescuer blows, we have to really get it right.

What did Stephen Covey say? “To be, and not to seem...”

Now I work in conservation finance, building infrastructure that contracts with Indigenous and local communities directly, without intermediaries. It is sometimes dismissed as utopian. What I have learned, three platforms deep, is that there is nothing utopian about it. Every layer you remove between two parties is a layer that cannot be used to extract from them. If you cannot insert a chatbot or a deniable employee in the middle, you cannot run the dark refund pattern on people.

Our companies have always run a concierge customer service hotline. As a founder, I have learned not to treat my clients like this. We run a human service. That means we don’t always follow through as well as a machine-enabled customer service line does, but we take our clients seriously and have always run businesses that treat them like people, not numbers.

The real defense

The real defense is doing something different

If you have a five-year-old refund file you never closed — or a soda-stained laptop you ate the cost of — you were not stupid, and you were not lazy. You correctly identified that the cost of continuing exceeded the expected return.

You might not get satisfaction from the people who wronged you. But strangely, you can definitely let the pattern stop with you.

And that in and of itself is satisfying.