1
$\begingroup$

I'm considering using R for my master thesis in order to do some analyses with the mvpart package. The problem is that I only have limited time to learn the program and to do the analyses. I already read through some information concerning R and I'm starting to get the feeling that it is impossible to get the know the program with the limited time I have.

I also have very limited background concerning statistics and statistical programs. I only did some limited work on SPSS and SAS. I have no background in programming.

Is anyone here familiar with R? What are your thoughts?

$\endgroup$

closed as not constructive by Andy W, whuber Feb 22 '12 at 14:02

As it currently stands, this question is not a good fit for our Q&A format. We expect answers to be supported by facts, references, or expertise, but this question will likely solicit debate, arguments, polling, or extended discussion. If you feel that this question can be improved and possibly reopened, visit the help center for guidance. If this question can be reworded to fit the rules in the help center, please edit the question.

  • $\begingroup$ You may find useful this post: stats.stackexchange.com/questions/138/resources-for-learning-r $\endgroup$ – Emer Feb 21 '12 at 14:32
  • 6
    $\begingroup$ If this question is about "How can I learn R in the next 5 days because my results are due in 6 days?", then it will be closed as off-topic (see our FAQ). If you wonder about learning R when coming from SPSS, you can have a look at the Quick-R website. Otherwise, please indicate what you are expecting from responses in this thread: "What are your thoughts?" is too vague. $\endgroup$ – chl Feb 21 '12 at 14:35
  • $\begingroup$ "R for SAS and SPSS Users" is another resource. $\endgroup$ – lockedoff Feb 21 '12 at 17:14
  • 1
    $\begingroup$ "A Handbook of Statistical Analyses Using R" by Everitt & Hothorn is good. It was used as the text in a couple of online courses I took (didn't see it in the link upstairs). $\endgroup$ – William Whitworth Feb 21 '12 at 20:09
  • $\begingroup$ You may also try R chat room on SO. $\endgroup$ – user88 Feb 22 '12 at 11:03
2
$\begingroup$

In general you can learn R as you go along especially if you only want to use it for a specific package. Look at the examples in the package and use these to learn how to apply the functionality to your own data. Quick-R will get you up to speed with some of the basic operations. Looking at source R code available in various packages will help you to devise your own bespoke functionality to, for example, visualise your results. I would say to go for it. You will have control over what you can do, you may discover some new and exciting results that would simply not be possible by using a closed source software. Learn by doing and not by reading is the way to go

$\endgroup$
  • 1
    $\begingroup$ If you hardly know about R, how do you want to understand th source code? Otherwise I agree with you. $\endgroup$ – user5644 Feb 21 '12 at 17:25
  • $\begingroup$ @lejohn For example, let’s say you see a nice graph that has been created in some R package with features you want to use but have no idea how they were achieved. Well, look at the source R code for this and see how it is done! This is pretty much how I learned R. I guess what I mean is that R can be challenging but when you look under the covers there is a lot there that you can use. Don’t be afraid to look! $\endgroup$ – martino Feb 21 '12 at 17:53
2
$\begingroup$

Does your university have any statistical consulting services? If you background in statistics in general is limited and you only have limited time, then learning R will not be enough, you also need someone who does understand the statistics enough to tell you what you might be missing. You may be able to do most of the work yourself, but you really need an expert who can look at your project (as early on as possible) and tell you if you are missing something important, or can confirm what you have done is sufficient for your question.

I have several non-statisticians that I work with who are savey enough to do a t-test or simple regression on their data, but they still like to run their projects past me. Most of the time what they did is fine, but they are happy to hear me confirm it. Sometimes there is something that is beyond their knowledge (how to deal with censoring, non-simple samples, adaptive designs, ...) and things are much better if I help them do the correct thing rather than let them do something that is simple but wrong for the situation.

That said, in the long run learing R will be useful for many uses, not just a single masters project.

$\endgroup$
2
$\begingroup$

If you are going to use a statistics program long-term, then it may pay to start investing in R knowledge now. But if you are just using it for this project, then I would "go with what you know".

$\endgroup$

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