On designing experiments

After meeting with my Co-PI and discussing the super cool experiment he wants me to do (here) it sounds like I’m in control here. Scary thought, right? This is that whole working WITH my Co-PI and not FOR my Co-PI, he trusts me to make good choices. I mean either way he’s going to need to sign off on it when I finish the design, but it’s kind of interesting to be the one who gets the say in how we do this.

Since I’ve been in the PhD program, I’ve done myself or helped with a small handful of experiments. Most of the time the design goes something like this, we meet with everyone involved, discuss the things we would like to find and how to find them, then the senior people in the group lay out how they want the experiment to go. It’s not that I don’t get any input or that I can’t suggest changes, it’s just that the design itself is basically pre-made by them and I’m just suggesting minor changes to make it work better. Most of the time it’s timing of things or reminding everyone that we need to make sure the experiment isn’t too long. An hour is best, two tops because setup takes an hour and we don’t want the person to be bored.

This is very different. My Co-PI and I discussed what we could find from the basic experiment we set up. He highlighted a few of the things he was thinking about, then asked me to come up with what I want to look for. As he pointed out, we have a lot of variables we could play with. So I could in theory do a LARGE (emphasis on large) experiment or I could do a medium experiment. The differences being the time to do the experiment and the conclusions we can draw from the experiment.

He want’s me to make sure that the data we collect will help me with my PhD progress and help us build on the R21 grant, which we should be hearing about soon. He’s also said that the goal if funded would be to roll it into an R01 grant. It was an odd thing to say since he was discussing moving to a different hospital (and state) sometime this year, but since we should be hearing back about the R21, if it gets funded I’m fairly certain that will weigh heavy in his decision to either go or stay (more here).

Anyway, getting off topic (per my usual style of writing). Basically, now I get to design the experiment to my specifications based on the questions we want to answer. Designing experiments is difficult and this wouldn’t be the first time I’ve designed one, but it will be the first time we actually implement my design (assuming we don’t change much). Okay, not entirely true, some of the early work for my proof of concept “super secret” technique was from experiments I designed. This is different though and will lead to a publication so I can share it with the world!

The trick is isolating the thing(s) I’m interested in. When designing an experiment there are all these variables we can adjust or hold constant and depending on what we do the result and what we can conclude from the result varies. Despite the brevity of research papers, 12 pages at most depending on the journal and frankly given all the work that goes in that’s not long, there’s a lot of planning involved. Experiments require not only thinking about how to isolate the thing you’re interested in, but also minimizing other effects that could alter the result.

It’s a lot to think about and there will probably be some back and forth between my Co-PI and I about how to best isolate the thing we’re interested in finding. It always helps to get multiple opinions though because you’re likely not thinking of something or you may be missing something obvious all together. We plan on reviewing Monday or so the plan I come up with so we can get started right away.

The undergrad I will be working with is very new, as in a freshman, and I have no idea how they got into the program, but heck yeah good for them! They will need to be brought up to speed somewhat on all this and how it works. Usually we get med school students over the summer, so this is a change of pace. The fact that we will only be working with them for ~10 weeks means that we need to go fast. At the end of the internship they will be presenting a poster on the work they did with us and the hospital throws a mini conference for them to all showcase their work. There will be several dozen posters if this year is anything like the year before last. Last year, due to COVID, we had no interns.

Anyway lot’s of thinking ahead for me as we try to figure out the best (or at least most relevant) questions I want to answer when we start this experiment. I’m excited, this is going to be a great project and I’m feeling good about what we’re trying to accomplish. While there is still a lot to be worked out, it’s already looking to be a really solid experiment.

Maybe I’ll write out my thought process for actual experimental design, but that’s kind of like painting or writing code, the overall result is the same, but how you get to the end can be very different between people. In other words, it may not help knowing my process because what works for you may be completely different. I’ll give it some thought once I work through what I am going to propose for this experiment.