A few very respected scholars use SPSS...though they are old
How many of you are using R (instead of STATA, SPSS...), and why?
(52 posts) (2 voices)-
Posted 2 years ago #
-
Here's why you want to learn R if you can:
1) You will save your current or future employer thousands of dollars. Every upgrade of Stata now runs $500 or more.
2) You will save your current or future grad students, and their universities, 10's of thousands of dollars. See above and multiply for every grad student you send out into the world.
3) You will give your undergraduates capabilities that they can use in the real world, unlike SPSS which charges an obscene annual license and Stata with the handicapped student version
4) If you work with foreign students, grad or undergrad, you give them something they can actually afford
Posted 2 years ago # -
^ (1) most major universities have deals with Stata so that individual licenses are nowhere near $500 (intercooled is closer to about $100 for updates at my place)
(2) most people just pirate Stata anyway
(3) In the real world, most places use SAS (although some are adopting R)
Posted 2 years ago # -
R is preferable for analysis and graphics over any other option by a long shot. I actually use SPSS for data management, because it is difficult in R, I am very paranoid about **** certain things up when recoding variables or merging datasets, and SPSS makes it easy to do these tasks and check my work. SPSS is obscenely expensive only if you have qualms about torrented software, which I don't. It is stats software for kids, but has its uses.
Posted 2 years ago # -
^ I don't understand, if R is better, why do you still use SPSS?
Posted 2 years ago # -
^You do realize that some packages can be better for different purposes. For example, R is great at graphics and writing functions, and less good for data management.
Posted 2 years ago # -
"R is great at graphics and writing functions, and less good for data management."Although this might have been true a while ago, it just isn't anymore.
For big data you can use one of many extremely powerful relational database solutions (e.g. 'sqldf'), or try speedy, native indexing with something like 'data.table'. For fancy manipulation and reshaping, try the Wickham packages 'plyr' and 'reshape'.
R is not just good, it's great (!) for data management (but I may not know enough Stata to compare)
Posted 2 years ago # -
Can't you and R just get a room?
Posted 2 years ago # -
"not really" proves once again the value of reading comprehension.
1) I wrote that you'd save your students, dumb ****, most of whom will not end up teaching at "major universities."
2) Most of us have multiprocessor machines, and Stata/MP is $400 and up for upgrades
3) Most of us gave up pirating software when we got a real job. Grow up, junior.
4) SAS is used by data miners. I meant they could actually use to do statistical analysis, dummy.
Posted 2 years ago # -
Is there any way to convert a STATA program into something R can recognize/use? My biggest problem with switching to R is the fact that I don't know anyone else who uses it and if I'm working with a coauthor they'll almost always be working in STATA. Is there a way to make R run a STATA do file?
Posted 2 years ago # -
^ the foreign package allows you to read in STATA (and other) datasets. I'm not sure about stata do-files. It seems like there might be such a thing... I know that you can call C++ functions from R, but I'm not sure about STATA.
Posted 2 years ago # -
Is there any way to convert a STATA program into something R can recognize/use?
don't think a package exists to do this at the moment:
http://stackoverflow.com/questions/4166981/calling-stata-functions-from-r
But in terms of learning R, after familiarity with stata see
http://rwiki.sciviews.org/doku.php?id=getting-started:translations:stata2r
Posted 2 years ago # -
OP here.
Thanks for your posts. I'm new to stats, so I'll try and get into R.
I saw that Gary King from Harvard is using it, too, so it must be good! And yes, feel free to answer how stupid that remark is - I'm used to nasty comments by now :)
NinaPosted 2 years ago # -
Didn't Gary start all of the rest of us using it? Isn't that how methods advances in PS? Gary does something, and we all adopt it?
Posted 2 years ago # -
There were loads of people using R and S+ when Gary was still using Gauss.
Posted 2 years ago # -
Nina cannot be real.
Posted 2 years ago # -
I agree that Nina is not for real. No one can be that cheerfully naive and yet self aware at the same time.
Posted 2 years ago # -
OP here.
Thanks for all the posts. For those that don't like the puristic code writing in R, I just found http://www.rstudio.org/ and installed it. It has kind of a graphical interface, e.g. you can show the tables in an extra window so that they look like in excel (for me that's better than seeing a matrix in R directly).And the thing about me not being for real... I'm just a grad student from Europe who is not so privileged to be in a Department where all these questions are answered in workshops or by the supervisor. Thank God this forum is anonymous, otherwise I'd probably never ask these questions. And I can promise you, I'm not the only one with these questions out there.
Anyways, as always, thanks for answering my qestion and good night.
NinaPosted 2 years ago #