Advantages and disadvantages of statistical software

I saw lots of questions regarding choice of statistical software for data analysis. I want to choose some statistical software for data analysis.
We could use more than one statistical software for our purpose. When we look at common statistical software, we see three or four main programs. These are SAS, Mathlab, Statistica and R. Maybe Minitab and SPSS could be added to them. Some of them need high profiencies because they work via coding language. The non-code programs such as SPSS, have disadvantages because of their graphical results not improving and model intervention is limited. For this reason, using the program working with code such as SAS, R, Mathlab is very efficient for scientific studies. R is free statistical software and it has many options about data analysis. It is promoted by not only statistical packages but also other packages such as psychological and social approaches. I prefer the interpretion of my results via multivariate graphical methods which have cases and variables could be shown in the same graph (such as Cannonical Discriminant HEPlot, Principal Component Biplot etc.). These methods need heavy program codes but they are very effective for interpretion of our results. SAS is another choice but it is not free. Both of these software are strengthened and differentiated from other choices via packages and macros.


1 / 1  ·  6 Answers  ·  3918 Views

All Answers (6)

  • SAS I felt was good for qualitative analysis and SPSS is quiet strong with quantitative analysis. My supervisor in Australia once said to me Shazia Stats is like a bikini it exposes what you want to expose so use it intelligently :)
  • Joseph Tkacz · Health Analytics
    To be completely honest, each of the major statistical software packages (SAS, R, SPSS, Stata) are realistically pretty equal with regard to functionality (please don't kill me anyone!). Sure, specific aspects of one program may be better than another (e.g., graphics), but overall, these software packages were designed to essentially do the same thing: manage and analyze data. I use SPSS, which has a very powerful proprietary command syntax language. However, many researchers introduced to SPSS simply do not conduct the type of work that requires writing programming code for complex data cleaning and analyses, so SPSS rarely gets credited for these core capabilities. I have worked on projects with both SAS and Stata analysts, and have had no problems re-creating, in SPSS, anything that those programs could do. It really just comes down to personal taste and familiarity. R seems to be the most popular one these days as it is open source, and I have no doubt that it is likely just as sophisticated (although not as user friendly from what I hear) as any of the other programs available, including SPSS.
  • Craig Smeaton · University of St Andrews
    It really depends on your needs SPSS is a brilliant piece of software if you want to use predefined statistical test and you dont want to spend a lot of time preparing code and data sets. And on first glance it is similar to operate. R on the other hand allows a far wider scope of statistical evaluation, the software allows you to design and code your own test which in essence allows you to create your own tailor made statistical package for your project. R also allows models to be created and linked with other R features so that is a plus. On first use R does look difficult to use but after you know some key lines of code it is very easy to build upon this crating more advanced codes. In terms of what is better SPSS or R it depends on your field I believe SPSS is still the top software in the social sciences and is the best tool to teach statistics to all undergraduate students (physical & social science). But I do believe that at doctoral level researchers and above in the physical sciences and engineering should be using R or one of its equivalents
  • Mohammad Tahir · Sugar Crops Research Institute Mardan
    To me it entirely depends on the scope of the analysis in hand. Some analyses are best done using one software, while others done using some other specialized software. Anyone of software with which a researcher is at ease is good. However, R can be put at the top due to the reason that it is open source and is being contributed continually by researchers around the world.
  • Matthis Drolet · German Primate Center
    While the packages are all similar in their core functionality (many statistical techniques are used across the board), they do differ significantly with respect to less well-known and new techniques. On that front I find R to be the most innovative. Of course, that means you do need to delve deeper into the research of these ground-breaking techniques as not all of them will stand the test of replication. However, considering that all the base functionality is there, it is worth it to learn R and have that option later. And open source is a big plus! Yes, the learning curve is steep, but on that note I have two other important points:

    1) The learning curve is less semantic (as with other scripting languages) than conceptual. And learning the concepts involved in R can actually give you a much deeper understanding of statistics than most point and click stats programs can (yes, some offer both, but if you are going to learn scripting stats, I would still recommend R for the other reasons above).

    2) I highly recommend Rstudio to get you started. It gets a lot of the hassle of learning R out of the way at the beginning and there are enough tutorials on learning R online with Rstudio now, which makes it even easier. And it is the one GUI that I find really works for R without taking away the stats scripting that makes R so powerful.

    ps. no, I am neither an R or Rstudio developer, just a happy user ;)
  • Fatih Kahrıman · Çanakkale Onsekiz Mart Üniversitesi
    Dear Drolet
    I agree with you about advantages of R. First of all it is free and than it is very innovative. I am neither statistician nor R developer too. But I see clearly that R offers many approches for special options researchers like me. Except of SAS and Mathlab, other programs could not solve the problems about special subjects. I think so that package developers are also employees about special subjects. So if you study about a special research area you could clearly understand the value of software with command syntax language such as R and SAS.

    The last word as says the crowd "Best choice for you, ones who makes you want to."

    Best wishes

Question Followers (10)