- So usually rigorous science, or rigor in science is defined as being both robust and unbiased.

- Rigor means things at different scales. It means what you do on a daily basis. That means how you keep your lab notebook, the attention to detail of how you're setting up that experiment. But if you're not paying attention and being rigorous on a daily basis, you're compromising your ability to make progress on your project overall. At the other scale, rigor means how you collect a large body of experiments that you've done in the laboratory, and condense them effectively into a single protocol or collection of protocols that allow other investigators to follow that effectively as a script, that allows them to reproduce your results.

- I guess there are a couple components to doing rigorous science. Certainly, perhaps the most is providing to the field a result which is reproducible. And that comes from being rigorous about describing the methods you've used and controlling variables, et cetera. So most of the time in science, we don't actually make, or quite often, the right interpretation of the result. That may comes years down the line but if everybody, if the result is because it was done, experiments were done rigorously, as a valid result, people can go back and use that result to make informed models and decisions. If it's a fake result, because it wasn't done rigorously, you can mislead the field through generations.

- I think this issue extends all the way from a legal situation, where someone knowingly falsified data, to on the far other extreme, recognizing the imperfection of science itself. Of course, in the middle of there, of this range is, have I done the experiment as well as I possibly could?

- So obviously, you want all experiments that you do in your hands, under your conditions, to be reproducible. So that is an absolute must. There's inherent noise in biological systems that is going to be inherent variability in biological systems and so you want to make sure that any effect that you're seeing is not a result of that inherent variability but as an actual true effect.

- When I think about reproducibility, I think of it as two levels. The first one is, is it reproducible in my lab and the second one is it reproducible to a community? What I request from people, in the lab, to make sure that somebody else in lab has seen, has looked at and seen what they are reporting. Once a person is sure about a result, I ask them to show that result to somebody else, to get that other person's opinion. That's one. The second thing that I do is that for every paper that we publish, I have personally looked at that data. And I'm not talking about pictures, I have actually looked at the strains. So I have my own interpretation of the phenotypes, usually for all figures. And then the third thing is that as soon as the result is ready for publication, I encourage people in the field to look at it. I pride myself on running a collaborative lab. We aspire to be generous with our reagents and our knowledge, we engage other people in the community and make sure that they benefit from our knowledge but also they see what we're looking at and they challenge it.

- I think it's a much more effective argument to be having a conversation about openness in science, than rigorousness in science. When we talk about rigorous experimentation, I think what the NIH would like to say is that, if someone did that same experiment, as a jumping off point, to follow up, that they should get the same result. But then that also means that we also have to acknowledge that there's a precision to doing experiments that doesn't always get translated to say, a protocol or even worse, how protocols get written up for papers. And so what that might mean is that there might need to be an openness between labs to discuss the specific situations under which experiments are done so that precision can be maintained.

- And obviously you want that. You want other people to be able to reproduce your work. In fact, it's a tribute to you. If you come up with something that's truly interesting and important, then the greatest tribute is that other people will actually want to reproduce your experiment. The most valuable commodity you have as a scientist, is your reputation. That is the single most, the greatest value that lasts with you, whether you're a graduate student, a postdoc or a senior investigator. So your credibility as a scientist really demands that you look at every experiment, not through the lens of what was easy to do but what was best, actually, for ensuring that the way you did that experiment is actually giving you a correct answer.