Learning and applying statistics in med school and research

[Bovary]

I have a surface-level knowledge of statistics, so I'd like to teach myself a few things about practically applying statistical methods to data sets in the next ~7 months before starting med school. I think this would be a useful skill to have as I re-enter the lab after several years of not doing research, and I rather learn the skills now when I have ample free time than once school starts.

Is it worthwhile to learn to use R or another statistical software?
Does the biomedical field favor one software/languge over others?

Has anyone used helpful (and ideally, but not necessarily, free) online resources/courses for this?

---

Members don't see this ad.

 

[Pholaffle]

Tried to help a friend use R in one of her ecology labs and I cringe every time I hear about it. My research labs all used PRISM graph pad which I believe is much more user friendly.

 

[mimelim]

Most people use STATA or SAS as @SouthernSurgeon points out. However, my guess is that if you are going to learn something well, R would be the choice. It is the standard in virtually everything outside of medicine and slowly it will take over. I have two people that I run my stats by, one is a statistics grad student at Rice, the other is applied math at Harvard. They both laughed at me when I talked about using anything but R.

My guess is that Medicine is a late adopter of R because they actually have money to buy the packages and/or they just pay people to do it for them.

ps. We are going to launch our own R statistical package dedicated to tools for surgical research sometime later this year 🙂. User friendly++

 

[bluejay456]

To be honest, unless you plan to be the person doing all the data analysis don't worry too much about what software package you choose. If you have a basic understanding of how programming works you can apply your knowledge to any package. I just picked up STATA after taking a previous course in MATLAB and it was insanely easy, you can't even call it scripting lol. If you're feeling adventurous maybe look into learning Java. Ive heard the programming skills you learn with that language is really generalizable to a lot of different languages

 

[WillburCobb]

Learn R. As mentioned above it won't be long until it becomes the dominant stats software in all fields - plus it's free and easy to learn. Andy Fields' Discovering Statistics Using R is solid (he has the same books for SAS and SPSS if you'd rather go that route). He has tons of free supplements and data sets on his website that you can mess around with as well. It's geared towards behavioral sciences, but he covers all the theory behind everything and its very applicable towards most other fields that use stats (i.e., it'll cover what you need to know for biostats and medical research).