Stats & Physics Blend: Pursuing a Career in Research

[quanto]

What sort of a blend would this be? Let's say I am interested in being part of a research in some physics discipline but not necessarily as a physicist but as a statistician/data analyst. Like The guys/girls who compile, analyze etc the data from experiments on particle interaction, for example. And they pass this data on to the scientists to interpret and analyze it in their own terms. So on the same team but a different role. Is there really such a thing? How would one go about designing their undergrad/graduate studies to go about pursuing such a career?

 

[fasterthanjoao]

I'll assume that studying physics is important to you as well, otherwise I might have suggested different undergraduate degrees.

Is there really such a thing?

Yep. Statisticians are needed (unfortunately for the community, not always present..) in one form or another in every research group. It is my feeling that good statistics is lacking in many research disciplines, though not highlighting physics in particular.

How would one go about designing their undergrad/graduate studies to go about pursuing such a career?

Pick up whatever statistics courses you can during your undergraduate. If you're looking to work in research, then a PhD in some sort of data analysis/statistics won't be hard to come by for a physics graduate - so you won't necessarily need any official statistics component of your degree. Even if you take a PhD in physics, you will be able to find something that involves analysing data sets, or you could find somewhere that will let you build a key statistics component into your research (you don't need to research in statistics, only apply it to the level that you can start to understand how all this stuff works). The 'statisticians' I know in my old physics department are just physicists with an interest in statistics - being the 'go to guy' for stats advice has the interesting effect where you will learn more about statistics, and so become even more-so that 'go to guy'.

 

[quanto]

cool advice faster, thanks!


i was also wondering if i want to work as a statistician in a research lab would i need to double major in that corresponding scientific discipline too? or is that something that the physicist, biologist can carry out on their own?

 

[Mute]

Where exactly are you thinking you want to do research? Industry or academia?

In academia, I would expect that, at least in physics, most graduate students are expected to know (or to learn) enough statistics to carry out their own analyses. They won't usually go searching for statisticians to analyze their data for them. Furthermore, a professor isn't going to hire a statistician to analyze the data, because they could use that money to pay a student or postdoc who knows the physics inside out and can do statistical analysis. If you want to do research in academia, in physics, you'd be better offer studying physics as a major and taking some stats classes, perhaps getting a minor in it.

As for other disciplines in academia, I'm not sure - maybe less-mathematically inclined biologists are more inclined to hire statisticians to do their analyses (or maybe they just all have statistics software like SPS to do it for them without hiring a statistician).

I don't know what the situation with research in industry might be.

Of course, you'll want to more thoroughly look into this, as this is anecdotal evidence I'm basing my conclusions on, but I don't expect them to be too wrong on average.

 

[twofish-quant]

What sort of a blend would this be? Let's say I am interested in being part of a research in some physics discipline but not necessarily as a physicist but as a statistician/data analyst.

There are some areas of experimental physics which the people involves are basically statisticians and data analysts.

Like The guys/girls who compile, analyze etc the data from experiments on particle interaction, for example. And they pass this data on to the scientists to interpret and analyze it in their own terms. So on the same team but a different role. Is there really such a thing?

In astrophysics and particle physics no. The problem is that in order to come up with something meaningful, you have to understand the instrument and the underlying theory.