The Experiments We Never Run

Somewhere, there’s a paper that will not get written. Before there’s a scientific paper, someone has to be there to ask the right question, design the right experiment, watch it fail, and try again. What happens when that person never gets the chance? Maybe they never made it to the laboratory, or maybe staying there required clearing too many obstacles that had nothing to do with being a good scientist.

As scientists, we spend so much energy worrying about what we’re missing in an experiment, adding controls, repeating runs, checking sample size and statistical power. But there’s a kind of missing data that we’re much worse at accounting for: the experiments that never happened because the person who would have run them never got the chance.

What are we actually selecting for

Becoming a scientist involves a lot of gatekeeping, and most of it is invisible until you’re the one being kept out. You get into a graduate program, find a lab, survive graduate school, publish papers, apply for fellowships and jobs, and keep convincing people at every step that you belong there. 

What I keep getting stuck on is who else gets filtered out along the way. If you can’t stand at a lab bench for six hours, does that say anything about whether you can design a good experiment to address a research question? If you need to run a protocol differently, are you worse at understanding the underlying biology? Being unable to travel to conferences, needing flexibility in when you work, needing information presented differently, none of that tells me much about whether someone thinks like a scientist. Yet, it absolutely affects whether they get to keep being one. If scientific training were an experiment, we’d probably call many of these conditions confounding variables. I’m not sure why we’re so much less bothered when the experiment is scientific training and the outcome is someone’s career.

Finding each other

I started DisabledInSTEM in 2020, right before entering graduate school. The idea was simple: I wanted disabled scientists to be able to find each other, as there were few around me to learn from. Since then, the community has grown more than I expected. 

Soon to enter its seventh year, the DisabledInSTEM Mentorship Program has connected hundreds of disabled and chronically ill scientists internationally.  Through this community, I’ve learned that mentorship between disabled scientists often gets much more specific than traditional career advice. How do I make this experiment more accessible? Has anyone adapted this technique before? Has anyone found light-filtering glasses that actually help when the lab lighting triggers a migraine?

Recently, a scientist in our community started a new job and ran into several accessibility problems almost immediately and turned to the community for help. The first issue was reading tiny lot numbers of antibody tubes. Their old lab let them zoom in with a phone camera, but the new laboratory doesn’t allow phones at the bench. They ended up using a magnifier and keeping it in the BSL-2 space. 

Another problem arose with flow cytometry and more community brainstorming was needed. The default cursor on the cytometer computers was a small white arrow, nearly invisible against a white background. They already knew how to make it bigger and change the color, but weren’t sure if they were allowed to touch settings on machines run by the flow core. So they asked. By the time they came back to use the computer, someone on the core had already changed it for them. 

I like sharing this story because nothing dramatic happened in it. There was a barrier in the laboratory, someone discussed what they needed, somebody else listened, and a setting got changed. We tend to talk about accessibility like it always requires new technology or major institutional overhaul, and sometimes it does, but sometimes it’s a cursor color or a darker background. Sometimes, it’s asking what would help instead of assuming what someone can or cannot do.

Not every problem is that easy, though. The same scientist also ran into trouble learning PBMC (peripheral blood mononuclear cell) isolation. As part of the protocol, you have to pipette off a thin layer of cells that’s hard to distinguish visually from the layers around it, and this time a handheld magnifier wasn’t enough since both hands were occupied holding the tube and pipette, often in a biosafety cabinet. Rather than struggle alone, they brought this situation to the DisabledInSTEM community. Would more contrast in the background help? Could you mark the tube ahead of time? Mount a magnifier inside the biosafety cabinet? People started throwing out ideas and working together to solve the problem, even though many had never done the protocol themselves. That’s one of the things I value about bringing disabled scientists together across fields and career stages. Someone may not know the protocol you’re struggling with, but they may have encountered a similar accessibility problem somewhere completely different in science.

What struck me was how ordinary the process felt. It was the same instinct scientists use whenever something in an experiment doesn’t work: identify the problem, change the conditions, and try again. A failed protocol doesn’t automatically mean the experiment can’t be done. But when disability enters the equation, the question can shift from How can we make this work? to Can this person actually do it? We are trained to modify experimental conditions when they prevent us from answering a scientific question. Why should we be reluctant to modify working conditions when they prevent a scientist from asking one?

Formal accommodations, accessible buildings, and institutional policies still matter. But no accommodation office can anticipate every strangely specific problem that comes up while actually doing science. Sometimes you need another scientist willing to look at the problem with you and ask, “Okay, what could we try?” That’s part of why I’ve come to think of the DisabledInSTEM Mentorship Program as an accessibility resource in its own right. Sometimes mentorship is helping someone choose the right lab. Sometimes it’s helping someone figure out how to see the cells.

The people we can’t find

Six years after starting DisabledInSTEM, what worries me most isn’t the people we’ve already reached, but it’s the ones we haven’t. Most of our growth has come through word of mouth and social media. Someone may find us, tell a labmate about us, share a post, or point someone struggling toward the mentorship program. The obvious problem with that model is that people have to find us first. 

I worry about the student who runs into accessibility problems and assumes the problem is them rather than the way the lab was designed. Or the scientist who has no idea someone else has already solved the exact accessibility problem they’re stuck on. I think about the person who’s never met another disabled scientist further along and starts quietly shrinking what they think is possible for themselves. We don’t even know how many disabled scientists are out there to reach, because plenty of people won’t identify publicly, or at all, particularly when disclosure can feel like a risk rather than a neutral fact. Unfortunately, I don’t have a fix for that. Some people will eventually find their way to a community like DisabledInSTEM. Some will have a mentor or colleague who helps when problems come up. Some will build their own workarounds and stay. And some will slip through the system, and I don’t know how often we even notice when that happens. The difficult part is that we don’t get to know what disappears with them. 

I don’t think accessibility means lowering the bar for who gets to be a scientist. I think it means being a lot more honest about where we’ve put the bar and what we’re actually measuring with it. Science is supposed to be hard. Experiments fail, hypotheses are wrong, papers get rejected, and good scientists spend a ridiculous amount of time confused. That’s part of the job. Having to reinvent a technique from scratch because nobody thought about whether you could access it is not. Having to solve the same accessibility problem alone when someone else has already figured it out isn’t. We shouldn’t mistake either one for rigor.

Six years after starting DisabledInSTEM, the question I keep returning to isn’t only how we support the disabled scientists we’ve found. It’s how we reach the ones we haven’t. We can’t search PubMed for the paper that was never written. We can’t know what disappeared when someone left science before they had the chance to ask their question. Somewhere, there’s still a paper that will never get written. Somewhere before that, there’s an experiment nobody got the chance to run.