4
$\begingroup$

I'm completing my undergrad this year and I've finished my graduate applications. I've been accepted to Oklahoma State University MS Business Analytics and Data Science. Looking through the curriculum I noticed that OSU is heavily focused on SAS.

Part of their core courses are Descripting/Predictive/Advanced Business Analytics that include a lot of SAS coursework. You can also earn SAS Academic Specializations through the coursework and OSU is a SAS partner.

I know SAS is an old, expensive program that many large, private business use. Especially in the healthcare/medical industry. I'm not targeting a specific company/industry that uses SAS so I'm wondering if this is going to be a disservice for me.

Would it be worth it to pass on OSU's program because of their heavy SAS focus?

For reference I've also been accepted to BU MS Applied Data Analytics and Rockhurst MS Data Analytics and none of their curriculum includes SAS.

$\endgroup$
2
  • $\begingroup$ Related and possibly informative: R vs SAS, why is SAS preferred by private companies? $\endgroup$ Sep 16 at 6:19
  • 2
    $\begingroup$ Although while this is open, I've added an opinion, I fear that it is too inviting of opinions to be a good question here. Also, inviting comments on specific institutions or degree program[me]s is not a good precedent. $\endgroup$
    – Nick Cox
    Sep 16 at 10:00
6
$\begingroup$

No single technology guarantees you the job or can be considered a life-long career investment. Different companies or institutions use different technology stacks, the trends also change over time. The fact that Python is popular nowadays does not guarantee that it will be popular in five years. Choose the program based on the skills they learn, not the technology. You can always learn the technology later on.

For technology, it is useful to know some programming language, as it makes learning other languages easier when needed, and knowing SQL in data-related professions is always useful.

The image below can serve as a proof that you would probably never be able to learn or use every technology that was created and someone ever found useful.

Crazy number of logos of different big data technologies.

$\endgroup$
2
  • 3
    $\begingroup$ (+1) Not knowing a specific technology is an easily fillable skill gap & prospective employers - or at least the savvy ones - are aware of this. Plenty of courses to send someone on who ticks all the other boxes. $\endgroup$ Sep 16 at 8:47
  • 5
    $\begingroup$ (+1) I started with Fortran (then usually called FORTRAN) in 1973. The existing statistical software available locally was just not programmable enough for what I needed. I stopped using Fortran not many years later, but it was a very good place to start. Over a career you can and should expect to have to change software many times -- or even to want to learn or play with different software some of the time. $\endgroup$
    – Nick Cox
    Sep 16 at 9:58
4
$\begingroup$

Yes, SAS is an older statistical programming language. This is not necessarily a bad thing; a strong data scientist or statistician will be good at their craft whichever language they prefer. Plus the FDA, pharmaceutical industry, and many academic researchers actually prefer SAS. However I would strongly encourage you to learn other languages like Python and maybe R - you'll be more marketable when you graduate.

Without comparing their programs, tuition, etc. I couldn't tell you whether you ought to stick with OSU or enroll with one of the other universities. OSU's SAS focus is a concern - they should be offering courses in other languages as well. Perhaps you could take computer science classes in Python on the side at OSU?

I've worked in public health and the healthcare industry for many years and am currently employed with Cigna as a statistician. Most of the data scientists use Python at Cigna, plus there are some statisticians like myself who prefer R. A few people still use SAS, including many valued employees. I learned SAS in graduate school 20 years ago myself, but rarely use it now. Cigna is very accommodating to your programming preferences, but we're definitely shifting towards more Python programming.

If you have the chance, also learn some basic SQL programming, which we use for extracting and manipulating data in Amazon Redshift or Hive on our Hadoop servers (although we're transitioning away from Hadoop to Apache Spark at the moment). SQL + Python coding skills is a great combination. And familiarize yourself with GitHub and GitLab for sharing code with colleagues - another valuable skill.

Hope this helps, good luck!

$\endgroup$
8
  • $\begingroup$ Thanks for the information, this is helpful! OSU isn't entirely SAS they do cover other languages like Python and R and visualization software. Compare that to the other programs like Rockhurst and BU, they don't have any SAS. Makes me wonder if I would be missing something at OSU that I would be able to get somewhere else that doesn't have any SAS. $\endgroup$
    – MyNameHere
    Sep 16 at 4:12
  • 2
    $\begingroup$ @RobertF is not correct that FDA prefers SAS, and pharma is using R more and more. But to the original question, as of 2021 a program that concentrates on R and Python is more in tune with what's happening, unless business analytics is completely divorced from data science (I don't think it is). If the OSU program is more the 1/2 SAS I'd be concerned. And also there is a correlation between software choices and out-of-date statistical methods choices. $\endgroup$ Sep 16 at 11:19
  • $\begingroup$ @FrankHarrell Thank you for the correction. Yes - R is also accepted by the FDA, don't know about Python. $\endgroup$
    – RobertF
    Sep 16 at 12:48
  • $\begingroup$ FDA is software agnostic, and some of the big pharma companies now have major R initiatives. $\endgroup$ Sep 16 at 21:47
  • $\begingroup$ @FrankHarrell They were initially a little wary of young, open source languages, weren't they? $\endgroup$
    – RobertF
    Sep 17 at 11:58

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