MD [...]

[fa21212]

---

Members don't see this ad.

[...]

 

[wholeheartedly]

Does anyone have any experience with SAS or R? Any recommendations on which one to use for statistical analysis on large clinical data sets? I prefer to use Mac OS, but can use Windows on a virtual machine if I need to. Also, I’m new to the statistics game if you have any recommendations with regard to getting up to speed on either programs I would greatly appreciate it.

really depends. I've used both. SAS has a pretty steep learning curve, R is a bit more intuitive. SAS is a bit better to handle really really large datasets, but not sure you'll hit that size. SAS is often considered foundational at a lot of places, so there might be more institutional support (biostats peeps, user groups, etc.). Honestly, I'd pick whatever you research group or collaborators have experience with. That way you have easy access to experienced users if you get stuck.

I ran SAS on my macbook pro using VMware Fusion, got me through grad school but it did strain my laptop a bit so make sure you have enough RAM and otherwise high performing enough system (more than SAS says you need). There's also a free university edition that gets you a SAS version that lets you do a lot of things, additional training, and works on mac or pc. It is more limited than the education version you purchase through school, but depending on what you need that might not be an issue.

SAS resources- SAS website, UCLA SAS site, The Little SAS Book, and I really recommend SAS for Epidemiologists by Charles DiMaggio which is essentially book form of a semester long SAS class and available for download as an ebook. can't remember what I paid for it, but it was well worth it. there's also SAS classes on sites like Udemy, coursera, edX etc.

R was more intuitive to me, but I didn't spend as much time learning it since grad school emphasized SAS. R runs more like a programming language, so you can get it to do other stuff like run programs on your computer. One group used it to run a program to automatically pull QC data off their clinical lab chem analyzer and analyze it. Quite a few places are moving to R.

R- most of my learning in R has been conference type settings and web resources, but there are R books on amazon and free online courses on the same edX and coursera type sites.

 

[71ubpmk]

I also prefer SAS, and the free university edition is awesome, but you may need the full version depending on how large your data set is. If you don’t have experience coding, it may be helpful to look into Stata, which is completely menu-driven.


Sent from my iPhone using SDN