Wow, n=3 independent experiments finally makes sense…

3 December – really??! Where did this year go? I don’t know the answer to this question, but I certainly look forward to Christmas. It has been an extremely busy but exciting year. My PhD project seems to have gained momentum, the models I investigate are more or less established, and I enjoy what I am doing. Can’t help the feeling of insecurity even when I am writing this; the recurrent question: “What if it goes wrong just because you state that it currently goes well?” Rubbish thought, dismiss it. Need to believe in myself, but it has proven more difficult than anticipated..

One of my main discoveries this past year is that n=3 independent experiments really matters. Let me explain. In biomedical science, particular in experiments involving cells, you usually repeat something at least 3 times on different days – ideally also on different cells.. This way, if you observe the same result with every repeat, you can be more confident that your findings are robust. We tend to laugh at this because from a statistical point of view n=3 is by no means a magical number.. Although scientists love to perform statistical tests on n=3, from a statistical point of view it doesn’t make much sense. So when I first entered the field, I thought that n=3 was quite a peculiar idea because it remains statistically irrelevant.

Statistically irrelevant doesn’t mean biologically irrelevant, though. Science is expensive, so performing the same experiment n=10 is unlikely to happen because your finances will suffer. Nonetheless, performing it minimum 3 times really is a good idea as I have discovered following endless worries that I might have messed something up. Perhaps a consequence of my pathological perfectionism, I often myself doubting every single thing I do and whether I might have messed something up without noticing? It doesn’t help that I spend most of my wake hours culturing cells, which at some point becomes quite automated so you just do stuff; yet, what if my autopilot has failed me without me noticing? Is everything I do wrong? Disaster. Rubbish thought, dismiss again. You see how easy it is to enter a vicious cycle of worries and self-doubt? Thankfully, I need to repeat it anyway, right? So hopefully, even if my autopilot might have failed me at some point, it won’t do so in a subsequent repeat. So if n=2 is different from n=1, n=3 surely will either conform to n=2 or n=1, if not – go for n=4 and n=5!

I wonder if others come across similar doubts and worries? Do share your comments if that’s the case. It usually helps to discover that you are not alone! Just remember, n=3 is there for a reason..

slide1