Can you work in physics research with a data science degree?

Question

I just finished my first year studying physics but honestly I didn't like it at all. I started studying a day before every exam and skipped almost every class. However, I love physics, I read books at home, watch documentaries and research on my own. I think it is the most beautiful field of human knowledge, but I hate the way it is teached at university and I really prefer experimental physics over theoretical physics.

I'm also very interested on AI and data analysis, so I'm thinking of switching careers to data science.

Would it be possible to work in physics research with a degree on data science? (With physics reasearch I mean doing data analysis of experiments and observations in a research group, for example to map galaxies, to classify particle crashes from a particle accelerator, etc.)

I think studying data science may be the best option for me because with the marks I have on physics I won't be able to do research. But I think I will be able to excel at data science, I'm studying every day on my own and taking online course after online course.

So what do you think? I don't want to continue with the physics degree, but I don't want to “say goodbye” to physics forever. Any advice is appreciated.

Answer 1

However, I love physics, I read books at home, watch documentaries and research on my own.

Engaging with popular books and documentaries on physics can be exciting and inspiring, but it's almost completely unlike actually performing scientific research. Actually doing the research involves a lot of inescapable drudgery: obsessively going over complex chains of mathematical reasoning, debugging huge and perhaps obscurely written computer programs, hours spent aligning components on an optics table, etc. All stuff that has to be done, but may not provide the easy thrill that comes from reading about 'big ideas'.

A friend studying geology went on two research cruises in the south Pacific in their first two years of graduate school. Exciting! Exotic ports! Big picture conversations about geology in the mess! The next three years were spent in a sub-basement, looking through a microscope, picking out little bits of volcanic glass from crushed rock samples using tiny watercolor paintbrushes. Nature's secrets are frequently hidden behind curtains of unbearable tedium. I don't intend to be discouraging, but you can't just blow off the hard, boring bits of a subject.

Would it be possible to work in physics research with a degree on data science?

Maybe? It wouldn't be an easy path unless you turn out to be an absolute star in data science. Most physics laboratories just use their grad students and post-docs do all the programming and data science work. As @anyon mentions in a comment, some of the very biggest centers like CERN and NASA do hire data science specialists. However, counting on that means narrowing your choices to a very few employers and geographic locations, and they still might expect you to have a significant formal background in physics or astronomy. It's my experience that the life sciences labs are bigger employers of data science and software staff than physics labs.

For the moment at least, data science does provide a lot of opportunities. If you find you like data science work, it might be a good choice. With some luck and flexibility you might eventually be able to land a spot in a physics research team. Just be aware that you are aiming at a tiny niche, and be prepared to pivot in other directions.

Answer 2

My close friend, a professor at a highly respected university, is in charge of a physics lab doing nano- and biotechnologies. Having a wealth of experimental results to process, he is now seeking to hire a postdoc or a highly advanced PhD student proficient in big data. However, the professor has firmly stated that a candidate should have taken the intermediate-level courses in physics back in their undergraduate years. It doesn't matter for this professor if a candidate has a BS degree in physics or chemistry or engineering -- but (s)he absolutely must be educated in physics at the intermediate college level, at least.

This story is, of course, just one datapoint -- but, trust me, a representative one.

Answer 3

Yes (with reservations)

I would refer you primarily to this great answer on the topic of the "research computing" career path

https://academia.stackexchange.com/a/178616/173832

Many national labs and universities have dedicated groups of programmers that play a critical role in scientific research. This is more like being a techinician/lab worker than a post-doc. You write code for researchers, or help them improve their existing codebases and workflow.

This is becoming an increasingly common and recognized career path, see e.g. the The United States Research Software Engineer Association

Source: I work as a data scientist/scientific programmer in a small earth science research group

Answer 4

No

I've spent around 12 years in university studying/working in physics and if you didn't study it, there's no way you'll ever work in it professionally. It is an insanely hard subject. Physics requires advanced mathematical skills and builds strongly on top of its own concepts, like Maths. You will have no success for example in electrodynamics, if you didn't already study classical mechanics for basic concepts like Lagrangians and Hamiltonians, let alone all the math required for doing calculus on vector fields, green functions, etc.

Frontline physics is HARD, for the very sole reason that if it were easy, it would have been done already. For example, data science would come in very handy in particle accelerators where you need to correlate datapoints quickly. But to get a basic understanding of relativistic scattering physics reallistically takes a masters degree in physics (5 years studying) and that's the "easy" textbook examples. If you already knew you'd want to focus on that, it could probably be done in 3, if you gave it all.

Physics is a really competetive field, with lots of PhD students and less post doc let alone permanent positions. Not to take away from data science, but at the end of the day, it's math, and physicists are good at applied math. I don't see any reason why a PI would choose a non-physicist over a physicist.