My research institute’s director submitted five years of data I collected — 200 fieldwork days, my own server, my own protocol — to the EPA to win a $2.1 million contract, and I found out when an EPA program officer called to congratulate me on a partnership I hadn’t agreed to.

My research institute’s director submitted five years of data I collected — 200 fieldwork days, my own server, my own protocol — to the EPA to win a $2.1 million contract, and I found out when an EPA program officer called to congratulate me on a partnership I hadn’t agreed to.

My name is Dr. Miriam Fong. I am a marine biologist. I spent five years collecting the dataset Raymond Stokes submitted to the EPA. It is on a server I administer. The IRB names me as principal investigator and specifies that data retention rights belong to the PI. Raymond signed the IRB. His signature is at the bottom. He approved my ownership and then used the data as his.

We were on the water when the first crack appeared, though I wouldn’t know about it until I was back on land. It was a Tuesday morning in October, the air sharp and smelling of salt and diesel exhaust. We were on a decommissioned research vessel I had negotiated special access for, a sixty-foot aluminum hull that cut through the chop of the bay with a heavy, unyielding grace. I was running a day trip to take water samples at three established monitoring stations. I stood on the aft deck with a young research assistant named Kevin, directing the sample collection protocol.

“Depth at exactly four meters,” I told him, watching the cable spool out into the gray-green water. “Not three point eight. Not four point two. We need the salinity controls to be absolute. Label sequence goes immediately onto the glass, then the chain of custody notation goes into the log.”

I did not explain this to him as an instruction to be blindly followed. I explained it as a philosophy of evidence. “We record it this way because in five years, when this data is compared to today’s readings, the collection methodology has to be reproducible. If it’s not reproducible, it’s not data. It’s an observation. The water changes every hour. The only thing that anchors it is the protocol.”

I have been making reproducible data for twelve years. Kevin nodded, his face serious inside the hood of his windbreaker, and wrote down exactly what I said in the waterproof logbook.

When I returned to the institute that afternoon, my boots were still damp and my hair was stiff with dried salt. I sat at my desk and began the upload process. My data files are on a server I personally administer — not the institute’s shared network. I set it up this way in my first year because I needed full control of the raw data integrity. File creation timestamps, version history, access logs — all mine. The IRB approval document for this longitudinal study explicitly named me as the PI and specified that data retention rights belonged to the PI. Raymond Stokes, the institute director, had signed that IRB.

My desk phone rang at 3:14 PM. It was an EPA program officer I knew professionally. He had attended a panel I spoke on at an international marine biology conference two years prior.

“Miriam,” he said, his voice bright with that specific, polished warmth of professional administration. “I wanted to congratulate you on the institute’s new contract. It looks like the Chesapeake longitudinal dataset is going to be put to good use. Raymond’s team submitted a strong application.”

I stopped typing. The cursor blinked on the server upload screen.

“Raymond’s team,” I said.

The program officer heard the pause. It was not a hesitation; it was an absolute cessation of momentum. He heard it, and the professional warmth evaporated from his voice, replaced by a sudden, careful neutrality.

“Was the dataset — I assumed you were involved,” he said.

“Thank you for calling,” I said.

I hung up the phone. I opened my email.

I sent Raymond a brief message asking to meet regarding the EPA application. I did not ask what the application was. I did not ask why I wasn’t on it. I simply asked to meet.

He responded within the hour. The speed of the reply told me he had been anticipating the email.

“Miriam, I’m glad you reached out,” he wrote. “The EPA application was a strategic institute-level decision — I should have included you in the process earlier, and I apologize for that. But the dataset was generated using institute resources, including the vessel and the lab facilities, which means the institute has a legitimate claim to it for contract purposes. I hope we can work through this collaboratively.”

He used the word collaboratively as a conclusion, not an opening. He was not inviting me to build something with him. He was informing me that the structure was already built, and that my required role was to reside quietly inside it.

I did not reply to his email. I opened my server instead. I exported the file creation timestamp history. Two hundred and fourteen data files. Creation dates spanning five years. Every single one registered to my administrator account. I pulled the physical copy of the IRB approval from my filing cabinet. I laid it flat on the desk. Raymond’s signature at the bottom. Principal Investigator: Dr. Miriam Fong. Data Retention: Rights reserved to the PI.

I logged into the EPA contract database and searched the recent awards. I found the application. My dataset was listed as the primary technical resource qualifying the institute for the $2.1 million contract. I was not listed as an investigator.

I opened a new browser tab and pulled the text of the federal regulations governing research data—specifically, 2 CFR Part 200, §200.315. I read the federal data ownership rule. The rule stipulates that data generated under a federally funded research project is owned by the institution. My data was not federally funded at the time of collection. It was collected under my own research agenda, using the institute’s vessel access as my only institutional resource. The federal institutional ownership rule did not apply. The IRB applied.

I designed the longitudinal dataset before I even joined the institute. I had a research agenda that required a stable monitoring location, and the institute’s geographic access was the only reason I took the position. I proposed the project in my first year. Raymond had approved the IRB and provided the vessel access. He had said, “This kind of long-term baseline data is exactly what the institute should be contributing.” He meant it. He just didn’t specify to himself, at the time, who would be the one contributing it.

By year three, the dataset was attracting real attention. I published a paper in Nature citing early findings. I was invited to present at the international marine biology conference where the EPA officer saw me. Raymond began including the dataset in his strategic presentations about the institute’s research portfolio. He referred to it constantly as “the institute’s longitudinal baseline work.”

I heard this phrasing during a staff meeting and noted it. That afternoon, I sent Raymond an email: “Quick note: the dataset is attributed to my research program per the IRB.”

Raymond replied thirty minutes later: “Absolutely — I’ll make sure to acknowledge your work appropriately in all materials.”

He did not acknowledge my work. He continued to use the phrase “the institute’s longitudinal baseline work.” I had saved that email exchange. It was the proof that he had been explicitly reminded of the data’s provenance, and that he had chosen to ignore it.

In year four, Raymond asked me to share the raw data files. “For a reporting requirement,” he said, claiming it was related to the institute’s annual federal reporting obligations. I provided a summary dataset—not the raw files—and noted in my email that the full dataset remained on my secure server per IRB protocols. Raymond said, “That works for now.” He submitted the EPA application six months later using my summary dataset, citing “access to the institute’s five-year longitudinal monitoring record.”

The EPA program officer who called me had assumed I was a listed investigator because he knew the data was mine. When I said “Raymond’s team,” he understood exactly what had happened. His call to congratulate me was, without either of us knowing it in the moment, the disclosure that prevented Raymond from ever being able to argue that I was informed of the application.

I sat at my desk with the EPA application open on my screen. I searched the entire document for my name. I found it exactly once. It was buried in a reference citation to my Nature paper, listed as a published source. I was not listed as an investigator. I was listed as a footnote.

I read the application’s description of the dataset again: “the institute’s unique five-year longitudinal marine monitoring record.”

I looked at the server window on my second monitor. Two hundred and fourteen files. My naming convention. My timestamps.

I closed the application. I opened the IRB approval. I found Raymond’s signature at the bottom of the page. I looked at it for a long time. The blue ink. The loop of the ‘R’. He had approved this. He had signed it.

Then I picked up my phone and called Constance Fisk, an attorney specializing in research misconduct.

I filed the ORI—Office of Research Integrity—complaint and the EPA Office of Inspector General complaint in the same week. The ORI investigates attribution and ownership of research data, not just authorship of publications. I did not tell Raymond I was filing. I sent a certified letter to the institute’s legal counsel informing them of the complaints, noting that I was the PI listed on the IRB, and stating formally that I did not consent to the continued use of my data in the EPA application.

I continued my fieldwork. The water did not care about the paperwork. I went to sea the following Thursday.

The ORI investigator meeting was held in a formal conference room at the ORI field office in Washington. The room had gray carpet, a long mahogany table, and the specific, heavy silence of federal oversight. I sat on one side with Constance Fisk. The ORI investigator sat at the head of the table. Raymond Stokes and his attorney sat across from us.

Raymond believed that institutional resources were institutional assets. He believed that if the vessel, the lab, and the institute’s geographic position enabled the data collection, the institute had a reasonable claim to the data. He had told himself this was the standard operating model for academic research. He was partially correct about the industry norm, and he was entirely, legally wrong about the standard that applies when a PI’s IRB specifies data retention rights to the individual.

Raymond’s attorney opened the discussion. He spoke with the smooth cadence of a man used to managing institutional liability. “The data was collected using institute resources—vessel access, laboratory facilities, and staff support. The institute’s standard intellectual property policy covers data generated using institute resources.”

Constance Fisk did not raise her voice. She simply opened her folder. “We have the IRB approval for this research program, signed by Dr. Stokes. The PI is listed as Dr. Miriam Fong. The data retention rights clause specifies rights belong to the PI. Dr. Stokes’s signature appears at the bottom of this document.”

She slid a copy of the IRB across the table.

“Additionally,” Constance continued, “the data was not generated under a federal grant—meaning 2 CFR Part 200, section 200.315, regarding institutional ownership provisions, does not apply. The institute’s standard IP policy does not supersede an IRB agreement that specifically addresses data retention rights.”

Raymond looked down at the document Constance had placed in front of him. He was looking at his own signature. He stared at the blue ink.

“I approved this in year one,” Raymond said. His voice was thinner than I had ever heard it. “I—I don’t think the data retention clause was the focus of my review.”

He looked at the investigator. He looked at the signature. He looked at the investigator again.

I leaned forward.

“I collected two hundred days of data on a server I administer,” I said. “The file creation timestamps go back four years before the EPA application was filed. The IRB you signed in year one names me as principal investigator and specifies that data retention rights belong to the PI. You signed it. You approved my data ownership and then submitted my data as the institute’s. The ORI investigator has both documents. They are in the same folder.”

Raymond’s attorney asked for a short break. Raymond stood up. He looked at the IRB approval, still resting on the table. He picked it up once, looked at his signature, and set it back down. He followed his attorney into the hallway. I watched him go. I had seen him walk through the institute’s corridors for five years, always with the confident stride of a man who owned the building. I had never watched him walk away from something before.

The ORI investigation confirmed that the IRB approval designated me as the PI with data retention rights. Raymond was found to have misrepresented the data’s provenance in the federal application. The EPA placed the contract on hold, forced a revision, and I was named as the Principal Investigator on the revised contract.

Raymond was given a formal research misconduct finding—a reprimand and a required data management policy revision at the institute. He was not removed as director. He continues in his role.

I am sitting in my office now. I still work in the same institute as the director who received a federal misconduct finding for misrepresenting my work. I see him in the hall three days a week. He nods. I nod. We have not spoken since the ORI meeting.

My field collection logbook for the next data collection season is on my desk. It is a waterproof, spiral-bound notebook. I carry one on every dive. It is the foundation of my methodology. I have opened the new logbook and written the first protocol entry in my own handwriting: the date, the collection sites, the methodology version number.

I have applied for a position at a different research institute. The application is currently under review. I do not know if they will make an offer. I do not know which institution I will be with when I fill the rest of the pages in this logbook.

But I am going to sea next month regardless of the outcome. The logbook goes with me. It always has. The institution does not own the logbook. It never did.

Raymond submitted my data as the institute’s because the vessel was the institute’s and the building was the institute’s. He forgot to re-read the IRB he had signed five years earlier. The IRB names me as the PI. The IRB specifies data retention rights to the PI. His signature is at the bottom. I did not put his name there. He put it there himself, in year one, when he was approving a research program that was mine. He approved it. Then he forgot what he had approved.