- Getting an experiment to work, that phrase, "to work" can mean many things, depending on who you're talking to. And actually, when I was studying as a faculty member, that was one of the hardest things for me to understand what the person meant when they used the words my experiment "did not work." The reason is that it can mean anything from I don't know how to do this procedure, because no one has told me and I have never done it before, to I just made the most important discovery of my life. And the reason there's so much space there is because this idea that the experiment worked sometimes is based on this conception that there's a right answer, which is your hypothesis, and it works when the hypothesis is proven correct. And I will argue that the experiment works when your controls work properly, and you can interpret the experimental result. - So when thinking about how to get an experiment working, so you know, there's that of different definitions of how does it to get an experiment to work. So, does working mean that you get a result that you're expecting? Does working mean that you just get to a conclusion that is interpretable? And so I think that is actually the goal of an experiment is to get to a conclusion that is interpretable. And so that usually requires having the appropriate controls. That usually involves quantifying all of the relevant data. That involves actually having the right comparisons to make, but I would consider it a working experiment, because you can interpret the results. - Now what the experimental tells you, if it goes in favor of your hypothesis or against your hypothesis, that is just, that's knowledge. If it's well designed, meaning that is has the controls, the positive control, the negative control, in place, the experiment worked. If the experiment did not work because the controls didn't work, because the technique didn't work, then actually that, that can be a frustrating process of troubleshooting, but to me, that is a tractable process of just getting the technique in place. Maybe there's some assumptions there that's wrong, maybe the machine broke, maybe some reagents need to be reordered, who knows. It depends on the actual experiment. But that is a tractable troubleshooting process. And if the experiment did not work because the controls worked, but the experimental finding goes against the hypothesis, sometimes I consider those to be the most exciting results. The ones that go against the way that I was thinking previously, as long as the controls are correct, as long as we are really, as long as we know that the experiment is actually telling us what's happening.