R has a programming language and is an software environment. It is by definition much more powerful than SPSS which you can only do few things in the GUI (and additionally you can set syntax, but it is no programming language).

For econometric analysis, you might have a look what it is possible when looking at the CRAN task view on 

http://cran.r-project.org/web/views/Econometrics.html

The learning curve is steep in R when you have no skills in programming at all. It is then a pain to learn R (or Matlab). But if you invest a lot of time, you can do more things. See also http://www.statmethods.net/about/learningcurve.html

Note that in R there are also GUI implementations, like 

library(Rcmdr)

I think quite comparable with SPPS's GUI.

R has a programming language and is an software environment. It is by definition much more powerful than SPSS which you can only do few things in the GUI (and additionally you can set syntax, but it is no programming language).

For econometric analysis, you might have a look what it is possible when looking at the CRAN task view on 

http://cran.r-project.org/web/views/Econometrics.html

The learning curve is steep in R when you have no skills in programming at all. It is then a pain to learn R (or Matlab). But if you invest a lot of time, you can do more things. See also http://www.statmethods.net/about/learningcurve.html

Note that in R there are also GUI implementations, like 

library(Rcmdr)

I think quite comparable with SPPS's GUI.

SPSS is more user friendly, go for that.

Hi!

As a programmer I already worked with some of the tools for economic modelling:

• Python. If you are planning to make models on hundred megabytes of of data you should use Python and corresponding libraries like "Pandas", "SciPy", "NumPy", "scikit-learn". iPython Notebook is quite MATLAB like GUI but faster, very cool stuff. Econometrics for Python: http://www.kevinsheppard.com/Python_for_Econometrics
• R. It is also good for the analyzing bigger data, like Python. It has many built-in features. but less freedom in programming. Here is a good description: http://cran.r-project.org/doc/contrib/Farnsworth-EconometricsInR.pdf
• MATLAB is still user friendly scripting language with a GUI and has many packages, good for research, but I have bad experience working with big data.
• SPSS is quite user friendly but it is limited to its feature. It would be good if you want to make some analysis, but I wouldn't do advanced research with that. It has a big overhead storing your data in memory, which may results big memory consumption.

Java. If you would like to work on gigabytes of data I would prefer Java, but it requires programming knowledge.

(Julia. I have no experience but many people say that it's worth to try it. http://julialang.org/ )

It depends on the size of data you want to work with. For smaller data (a couple of megabytes) I would use MATLAB, for bigger data I would I would invest to learn to use R or IPython Notebook. Interesting article: http://r4stats.com/articles/popularity/

Bests, David

It depends on your purpose. I do not know your interests and needs in terms of Econometric modelling. stata, OxMetrics, Eviews etc. are all good softwares. But if you decide to invest your time, R is really a good choice. It's powerful, flexible, cross-platform, free, and with many and many libraries soft. Also it has the large community of users, so you will not be alone.

I would generally agree with Roman Matkovskyy. There's no point in learning MATLAB if most of what you will be doing can be done fairly easily in Excel. Likewise, there's little point in using Excel if your lab or department or business provides you with a license for SPSS (which is basically Excel with a lot more ability to not understand what you are doing but report significant results because you plugged in numbers and SPSS said your p values were .01 or something).

In particular, if your choice is between MATLAB and R, the first thing to consider is that R is free and MATLAB is extremely expensive. I would only consider it if you either 1) have a lot of cash to throw around or 2) your department, lab, etc., has an institutional license you can use.

If you CAN use MATLAB (i.e., you can afford it or it is provided for you), then there are perhaps two major things to consider. First, R is a statistical language/software package. MATLAB can certainly run any statistical tests that R can, and has most of the statistical methods/tests/models that R has built in. However, it is designed more for computational sciences, engineering, and so forth. The name comes from "matrix laboratory", and there really isn't any software out there superior to MATLAB when it comes to exploiting the power of linear algebra with ease. Additionally, the GUIs available for R (RExcel, R Commander, RStudio, etc.) are fairly limited, especially when compared to MATLAB, which in addition to being the perfect balance between a programming/scripting environment has a large number of toolboxes/apps that run the gamut from a neural network tools and fuzzy logic tools to a slew of computational biology tools and of course an Econometrics Toolbox.

Also, because the environment is so much more user-friendly, you don't have to spend time e.g., writing the code for plots, then changing it and running the same plot again, then trying out another by writing more code, etc. Once you have your variables you can use graphical user interface plot functions (MATLAB has many).

However, MATLAB is again more a computational software program than a statistical one. As it is proprietary, there isn't nearly as much support. So even if you have access to MATLAB, if you are only or mostly concerned with statistical analysis, R is probably better.

A note, though, about Python. SAGE mathematics is a free-ware scientific computing software package that brings together many, many free scientific computing, statistical, and computational packages such as SciPy & numpy. However, it's a pain to use if your operating system is windows unless you know your way around emulators.

I am very thankful for all the researcher's answer and especially the comments and explanation given by Matthias Templ, Andrew Messing, Roman Matkovskyy  and Dávid Zibriczky helps me in understanding in which is better to learn as per my need. really there are pro and cons on every packages but still I feel as of now it is good for me to learn R language and to understand the library available. I welcome further comments regarding learning R. Thank you once again for all your valuable comments and valuable time spent over my question.