Take the 2-minute tour ×
Cross Validated is a question and answer site for people interested in statistics, machine learning, data analysis, data mining, and data visualization. It's 100% free, no registration required.

I am currently a PhD student in Biostatistics (finishing my 2nd year). My original goal when entering graduate school was to go into academia. Recently I have been debating this choice and wondering if perhaps industry will be a better fit.

I come from a strong programming background. I love using R because it is a lot more similar to the programming languages that I grew up with, and is even able to interface with them.

I abhor SAS. Whenever I am forced to use it, I feel like I am wearing 2 left shoes that are 3 sizes too small.

I know that SAS is predominant in industry however. Do you guys think I will just have to suck it up and learn SAS proficiently, or will I be able to get by with R?

share|improve this question
3  
4  
Any answer will depend strongly on the industry you're interested in. For pharma and other bio-related research and work, SAS has a very strong foothold, to the point that it is often implicitly assumed to be the software preferred by US regulatory agencies for submissions and reporting. (See this though!) However, in some other technical fields where heavy statistical analysis is done, you'll find R is currently (far) more prevalent. –  cardinal Mar 9 at 20:16
3  
I wish bio-industry would see the glorious light that is R :( –  Benjamin Mar 9 at 20:24
3  
It depends on what jobs you do in which industries. My suggestion would be simply to bite the bullet and learn it enough to be able to work with it moderately well (and to be able to put it on a resume), because even if you don't use for your analysis, you'll likely be working with/getting data from/giving information to people who do use it. It helps if you can understand the constraints under which they operate. –  Glen_b Mar 9 at 21:27
    
I abhor R. Whenever I am forced to use it, I feel like I am wearing 2 left shoes that are 3 sizes too small. I use Python. –  sashkello Mar 10 at 3:58

2 Answers 2

up vote 8 down vote accepted

you have to figure what is that you're interested in:

  1. programming statistics in R
  2. programming statistics
  3. statistics
  4. programming

If your answer is 2-4, then it shouldn't matter which language you use. if you already know R and don't want to learn SAS, then get certified in SAS. this will increase your chances of getting employed at places where they require SAS. once employed, you'll learn SAS and use it, and it wouldn't bother you at all.

Only if your answer is 1., you're in a tough spot regardless of the industry. I frankly don't understand personal attachments to particular languages and tools, especially for a young guy in the beginning of a career. it makes a sense for 60 year old dude who simply doesn't want to spend his precious time on learning something new when he can simply ride on what he already learned during his long life. learn SAS, Stata, Gauss, Python... you have plenty of time ahead.

share|improve this answer
    
Good advice! Thank you very much. I will take it to heart. –  Benjamin Mar 9 at 21:14
    
you're welcome. it took me 2 months to prepare for SAS Base exam, just daily 1-2 hours in the morning. it's a piece of cake. –  Aksakal Mar 9 at 23:33
7  
60 year-old dudes read this forum too.... –  Nick Cox Mar 9 at 23:35
1  
Note that FDA has hired a lot of statisticians who used R in graduate school. So now FDA uses R extensively. And more industry statisticians are using it - some in clinical and even more in pre-clinical. –  Frank Harrell Mar 10 at 0:04
    
@NickCox: some 60-year-old dudes...but I know at least one who doesn't and should! :) Aksakal's answer describes him pretty well, I'm afraid. –  Nick Stauner Mar 10 at 0:16

SAS is extremely expensive as an enterprise wide solution. It is used by some large organisations specially in banking and insurance. Many companies today are taking a different approach, looking for less expensive and scalable solutions. Open source is getting a lot of traction even in large organisations. I would start with R programming and maybe Python. Once you have enough exposure to these, and if you decide to learn SAS, it will be a lot easier to digest.

share|improve this answer
1  
software license costs usually are much smaller than the labor costs. if a company can't afford SAS license, will pay competitive salaries? i'd be concerned about this aspect. there are legitimate reasons to go with open source solutions and they are usually not the cost –  Aksakal Mar 10 at 0:40
1  
Paying for SAS is only part of the problem. Software development takes far longer than using modern tools, so more SAS programmers are needed per unit of production. –  Frank Harrell Mar 10 at 1:35
1  
@FrankHarrell, I don't think you can support the claim about unit of production. Any references? Anyone who tried to learn SAS and R knows that R is much harder to pick due to the total absence of product documentation. Its help files are garbage compared to SAS. –  Aksakal Mar 10 at 13:09
1  
I could not disagree more. There is an amazing amount of documentation about R - you just have to be good at filtering. I used SAS for 23 years so I feel qualified to compare the two languages and have witnessed first had how inefficient SAS programming is in the pharmaceutical industry. R is harder to learn in some ways because it is a complete programming language. –  Frank Harrell Mar 10 at 14:02
1  
@Aksakal: Licensing costs & vendor lock-in aren't trifling concerns. You don't buy SAS software in a shrink-wrapped box off the shelf in Dixon's - you need to cut a deal with SAS. So I shouldn't be concerned in the least about working for a company that I'd heard decided against a SAS licence on cost grounds. (Nor of course, for one that had decided it was worth it to them.) –  Scortchi Mar 10 at 22:04

Your Answer

 
discard

By posting your answer, you agree to the privacy policy and terms of service.

Not the answer you're looking for? Browse other questions tagged or ask your own question.