Developing intuition on top of technical skills in your research

Hey everyone! I’m sure there are a bunch of senior members here and I wanted to get your take on this. I’m currently going through the ANTS book (very slowly, since its something I do in the evenings) and I’ve been having a lot of fun with the implementation of the methods. Sometimes I feel I have too much fun in the implementation parts of my research—and tend to side line important things that would have more long term benefit, such as reading literature. I’m always a little bit worried I might become very technically skilled but lack the intuitive/creative part of my science. I’m at an early part of my PhD and am forcing myself to read literature, since I feel that really is the bulk of material which takes me from data to intuition.

I’d love to hear how the other members here have tried (or have achieved) to get a sweet spot in developing both technical and intuitive skills.

There are so many ways you could approach this. I can say something about myself and then make a more general comment.

I never had much of a big plan, particularly when I was a PhD student. I just worked a lot on whatever I found interesting and relevant. My life philosophy is to work hard on problems I find challenging and important (that is, make some kind of contribution), and I trust that good things will happen eventually. I do try to keep up with the literature, but (as you might imagine) my focus has been more on the methods.

Keep in mind that different people have different strengths. There are plenty of scientists out there who know the literature extremely well but aren’t so great at data analysis. You could build an entire career out of collaborating with people who have strengths that are complementary to yours.

2 Likes

Hi Asimio,
It sounds like you’re enjoying the methods, but want to get a better theoretical grasp. Hopefully your mentor and program can leverage your MA, comps, and dissertation projects to guide this theoretical development. Ultimately, it sounds like you hit the solution on the head here - there’s no substitution for reading the literature.

I’m personally a theory-driven researcher. So this ‘intuition’ is extremely important to me. Reading the big theory papers can help identify what the most exiting questions are. For me, this included Buzsaki’s Rhythms of the Brain book, and e-phys work by Karl Friston, Pascal Fries, Erol Basar, Wolf Singer, Thilo Womesldorf, Wolfgang Klimesch, etc. Also classic ERP work by Donchin, Polich, Luck’s book, Coles, Kutas.

If you know what theoretical areas of cognitive science you find most interesting (reward, emotion, perceptual decision making), you also need to go through that specific (non e-phys) lit. I guarantee you that eventually your combined knowledge of the topic + the technique will yield novel questions that other researchers haven’t thought to ask.

2 Likes

Thank you both for your answers. One question about reading papers beyond one’s current knowledge base: is the best way to read up (via easy sources like wikipedia) what you do not know? For example many times I find really interesting papers, but when I see equations and advanced statistical methods that I haven’t heard of, I get a little shaky. Sometimes (and often with very theoretical papers) I avoid reading papers because I do not have the background knowledge to understand them: I tell myself when I know these methods I’ll take a look. I’m beginning to think this is a bad approach, because I do not find myself returning to those papers. How have you approached this issue?

There are multiple levels of “understanding” a particular data analysis method. You might be able to understand conceptually without getting all the math, which is fine.

If you really want to understand the math, then you’ll need to learn the math. I realize that’s easy to write and harder to do. It takes time and commitment (and a lot of nights spent studying instead of watching Game of Thrones), but it’s possible. If it’s any consolation, I don’t have a math degree. I learned math because I wanted to understand methods I didn’t understand.

1 Like

Jump in feet-first. You can always re-read them again to get more out of them. I oftentimes skip math sections until I really want to sit down and chew on that particular issue.

1 Like

True, math is something I have to cover a prolonged period of time (I’ve been working on a habit so I can do a little bit every day along with some longer evenings)—no point holding off important papers since it’s impossible to be “done” with the math. I really think the ANTS book hits the perfect sweet-spot for most neuroscientists that have backgrounds in medicine, biology, etc. I just wish there were more books/courses with that in mind. Most of the students in my masters program grew a strong dislike for programming and theoretical neuroscience just because that middle ground was never found.

1 Like