SPSS suffers from the disadvantage that it has an annual license which is extremely expensive for an individual. So you risk teaching students to use software that they will not be able to afford to license. 

SPSS also has an unpleasant "in-app purchases" system where anytime you need to do anything sophisticated you have to buy add-ons. By contrast, Stata, for example, gives you the full analytic capabilities. Stata versions only differ in the size of the datasets they can handle and their ability to use multiple processors. That said, the limits on Stata/IC are 2.1 billion observations and 2,047 variables!

By contrast, R is harder to learn and use, but free. And Stata's student licensing scheme makes student licenses cheap and non-expiring. Stata also has a developing countries pricing system that makes a substantial discount off the regular price.

Excel's statistical methods have been widely criticised as inaccurate and lacking even basic safeguards against nonsense results (such as detecting perfectly collinear variables in a regression). It also does not teach a very valuable skill which is cleaning, verifying and documenting your data, because it has a very limited or non-existent capacity to do so. This means that an Excel datafile is not a proper way to archive data.

Check out SAS, who now have a "university edition" that is free. I can't vouch for it – SAS is quite a hard language, but a powerful brute – but it may well be their response to rapidly declining use in health research. SAS and SPSS are seeing use plunging catastrophically as Stata and R take over the market. 

Here's the link to the SAS University Edition: http://www.sas.com/en_us/software/university-edition.html

http://www.sas.com/en_us/software/university-edition.html

In my courses for undergraduates  I use Excel and Gretl. Excel is nice to 'visualize' things e.g. the product of two matrices or the transposed matrix. I use Gretl because it's freeware and all basic analysis can be done in there via click interface. So students can use this software at home and it's a little bit more intuitive than R. And if you want you can write programs in R and run them in Gretl.

Best regards

Hi, Good Morning

Your qeustion is very interesting one.  for undergraduate students Excel and SPSS are necessary and enough also.  Because, their knowledge and understandable are very less.  Another one reason is, they are having their speciality subjects burden also there na.  So, two softwares teaching is enough to them.

Thanks.

SPSS suffers from the disadvantage that it has an annual license which is extremely expensive for an individual. So you risk teaching students to use software that they will not be able to afford to license. 

SPSS also has an unpleasant "in-app purchases" system where anytime you need to do anything sophisticated you have to buy add-ons. By contrast, Stata, for example, gives you the full analytic capabilities. Stata versions only differ in the size of the datasets they can handle and their ability to use multiple processors. That said, the limits on Stata/IC are 2.1 billion observations and 2,047 variables!

By contrast, R is harder to learn and use, but free. And Stata's student licensing scheme makes student licenses cheap and non-expiring. Stata also has a developing countries pricing system that makes a substantial discount off the regular price.

Excel's statistical methods have been widely criticised as inaccurate and lacking even basic safeguards against nonsense results (such as detecting perfectly collinear variables in a regression). It also does not teach a very valuable skill which is cleaning, verifying and documenting your data, because it has a very limited or non-existent capacity to do so. This means that an Excel datafile is not a proper way to archive data.

Check out SAS, who now have a "university edition" that is free. I can't vouch for it – SAS is quite a hard language, but a powerful brute – but it may well be their response to rapidly declining use in health research. SAS and SPSS are seeing use plunging catastrophically as Stata and R take over the market. 

Here's the link to the SAS University Edition: http://www.sas.com/en_us/software/university-edition.html

http://www.sas.com/en_us/software/university-edition.html

In the past I have taught SPSS and Minitab to undergraduate psychology students. Both packages are good for this purpose. R is a difficult package to use, although several handbooks are now available to speed up the learning curve. It is probably the best package for those specialising in quantitative aspects of social science. If you like computer programming and know how to debug software, R is an excellent package.

In my opinion MS Excel and SPSS would be ok for the purpose. 

From my point of view I think to get familiarized with SPSS is good, because this program it’s easy and have a lot of funcionalities. Futhermore is very intuitive, if they learn how to run descriptive and inferencial statistics they would have no problems with this software.

I also think that there are special licenses for universities, so maiby that would not be a problem. In addition there are plenty of useful manuals of how to use SPSS and run different analysis.

R is so useful but hard to begin with it…

If you want to learn SPSS, there is a free book to download on my ResearchGate page.

It depends on which UG students you have in mind - majoring in statistics, biology, ... ? For statistics major students, R is definitely the best option in the long run. For others, probably EXCEL may be a good option BUT note that EXCEL could go wrong in getting a statistical analysis wrong! Other options like SPSS, MINITAB, ... however cost money! So I see no reason why students not majoring in statistics cannot be taught R - R Commander is particular is menu-based and could be easily learnt. 

I generally teach SPSS and Excel

Hi Eddie,

I would recommend R statistical software.

It's easy to use. And the good thing is that the software is small, then you can download packages depending on your field of interest. 

Also, the best part with R is that it's open source, so you don't have to worry about the license

Dear Dennis,

As Ronán put it, it is hard to learn R, one should read and thoroughly study over 100 pages. Most of my colleagues without help could learn and use it.

I like excel for undergraduate students.  Most students are petrified of calculating statistics and excel is easy for them to learn.  In most work environments it usually comes with the Microsoft package so most managers and and staff have easy access to Excel.  SAS, Stata or R should be used by graduate students who are going to perform more advanced statistical calculations.

Agreed on the above scholars' sharing on the advantages & disadvantages of various statistical software like R (free & versatility), SAS (free version for university students & all in one completeness) & SPSS (common but a bit expensive on software licensing) etc.

Aside cost & functionality etc, I observe many universities in Asia start exposing undergraduate students to AMOS & SmartPLS besides SPSS.  Also observe many universities in Asia still embracing more on quantitative research compare to universities in Europe, America and ANZ (in which more & more students are conducting qualitative & mixed method research).  SPSS for mainstream descriptive & inferential statistics.  AMOS (a module of SPSS) for covariance-based Structural Equation Modeling (SEM) and SmartPLS for Variance-based SEM.  Because nowadays the conceptual framework / research models are getting more complex than simple / multiple regression analysis in which AMOS & SmartPLS are coming very handy. 

I have become a Stata convert. I would recommend this over SAS b/c it's much more intuitive (to me anyway) and quicker to use. It has some pretty easy-to-use documentation tools (like variable notes and data file notes) that I still can't figure out in SAS. Essentially, it's designed for replicable data analysis that can be conducted by an individual researcher. Much cheaper and easier to manage licenses, installs, etc. than SAS. So unless you're training to be a government programmer specifically, I'd go for Stata over SAS. The obvious next option is R, which is great b/c it's free and highly flexible. It may be "too much" for some types of grad programs and students, but there are a lot of GUIs to ease the transition. 

Stay away from SPSS unless you already love it. It's limited in the types of models and survey weights you can use, and their customer support has collapsed after the merge with IBM. They seemed to have moved into the "big data"/analytics market more, so unless you're a major corporation, I wouldn't even bother. Graphics in Stata are as good or better than SPSS, and even SAS can now do nice output and graphs via ODS (which is turned on by default). I would go with SAS over SPSS, but Stata and R would be better choices b/c you can take them with you anywhere (doesn't require your employer to have a license like SAS does). If you're starting from scratch (or training students from scratch, I might actually recommend R because of its ultimate flexibility and the explosion of training options over the past 10 years).