Investigating scientific misconduct is hard—especially when your supervisor is an author

From ScienceMag:

I was 3 months into graduate school when I realized my project was doomed. I had set out to build on the work of a previous student, but as I ran into roadblocks, it became increasingly clear that the previously published work was fundamentally flawed. The data didn’t make sense; the results couldn’t be replicated. I raised my concerns with my supervisor, but he was convinced there was a reasonable explanation. I clung to his reassurances for a time, assuming no one would publish something blatantly wrong. As weeks of digging and hoping turned into months, though, the cracks started to widen. Eventually it was undeniable: The paper was riddled with serious problems.

My supervisor was the corresponding author on the published paper, and when I told him he grew indignant. “I don’t understand why you’re making such a big deal out of this. Mistakes happen all the time. People mislabel things, they forget. Do you really think that in 5 years you could look me in the eye and say you’re 100% confident in everything you did?” He leaned across the desk, his gaze stern, while discomfort washed over me.

“Yes,” I answered, with more defiance than certainty. Honest mistakes were one thing—mislabeling a tube, losing track of a sample— but to my fresh, first-year grad student eyes, to publish an entire paper built on a mountain of mistakes was inconceivable.

I had heard stories of sloppy science, and worse. But I saw them as cautionary tales, not something I would have to personally grapple with during my first year of grad school. For the most part, I had been taught to view science as a domain of rigor and diligence, kept on track by the guardrails of both scientific scruples and peer review. But I was beginning to realize the people who did science were just that—people.

To publish an entire paper built on a mountain of mistakes was inconceivable.

• Ph.D. student
• a research intensive university in North America

Because I couldn’t build on the work, my supervisor instructed me to redo the original publication, eager for me to show the problems were no more than minor oversights. I painstakingly repeated the methods, which involved reanalyzing data. It was frustrating to not be setting out in a new scientific direction. But eventually, after I completed my exhaustive retracing of the original paper, I had something to show for my work.

With a knot in my stomach, I carefully presented a list of issues and mistakes to my supervisor. I told him about incorrect data analysis and experimental design, results that couldn’t be replicated, and claims that were contradicted by the data. My supervisor’s initial patience and silence slowly gave way to defensive interruptions. He dismissed some discrepancies as minor and insisted other errors weren’t worth mentioning. Nobody likes the bearer of bad news, and I was aware I was risking my future. I’d need his support to continue in the program, and someday I would be turning to him for reference letters. But I couldn’t back out now, and I was resolute on wanting the scientific record to be corrected.

I pushed for a complete retraction of the original paper. My supervisor instead lobbied for a small correction, an addendum to gloss over the errors. As the corresponding author, he was in communication with the journal. The journal editors convened a special meeting and reached their verdict: The issues were too systemic and serious for a simple correction. The only viable course was to retract the original publication and replace it with a paper describing my analyses. I felt vindicated and relieved: The errors were as serious as I thought, and I had been right to expose them.

Afterward, other faculty members commended me for standing up for research integrity. Their support, however, couldn’t change the fact that I never wanted to have to choose between truth and peace again, especially not while existing under the thumb of my supervisor. I have maintained a good relationship with him through it all. But the experience was utterly exhausting and I understand why early-career researchers, if faced with a similar situation, might choose silence over speaking up. It’s not easy to point out errors, especially when they’re attached to the name of someone who holds great power over you.

Read More