0
$\begingroup$

I would be using such software to run multiple regressions using macroeconomics variables as independent variables to estimate other macroeconomic ones as single dependent variables.

I need to be able to run robust regressions including Weighted Least Squares(to resolve heteroskedasticity), Feasible Generalized Least Squares (to resolve autocorrelation of residuals). If using a regular regression, bypassing such models, would need to be able to run heteroskedastic resistant and/or autocorrelation resistant Standard Errors. Probably could also use having access to ARCH and GARCH models.

I am not a computer programmer, coder, Visual Basic expert. I am just a pretty good power user of regular Excel functions and its Add-ins programs. Thus, I am interested in something very user friendly with a menu or window interface that would not be too difficult to pick up.

Because of my ease of use requirements, I have no interest in SAS. Preliminarily, talking to a few economists they seem to gravitate towards EViews. I also hear good stuff about Stata. I sense EViews is a bit easier to use. But, Stata may be a bit more powerful. I hear SPSS is pretty good. But, I sense for econometrics models it probably would come as a distant second vs either EViews or Stata. I also hear a bit about Microfit.

Do you have any experience with those programs? Can you give me some insights regarding their respective trade offs? What would you recommend? Is there another program I should consider?

Improve this question
$\endgroup$
10
  • 5
    $\begingroup$ I recommend that if you are serious about econometrics modeling that you get over your phobia and learn R, SAS, SPSS, etc. To complain about programming is the equivalent of an accountant complaining of having to learn arithmetic. If you're doing your analysis with Excel, you're lacking in professionalism. There have been enough media stories about economists and pundits embarrassing themselves by (incorrectly) using Excel to do their analysis. $\endgroup$
    – rocinante
    Apr 2, 2014 at 23:36
  • 1
    $\begingroup$ @rocinante (I love the accountant analogy.) Real professionals use Excel and survive to tell the tale. :-) Seriously, professionalism is about the approach one takes to solving problems and mastering the skills needed to do that well. It's not really about what tools one uses (although you likely won't get any arguments about the potential pitfalls of using a spreadsheet). $\endgroup$
    – whuber ♦
    Apr 2, 2014 at 23:47
  • $\begingroup$ @whuber I'm not sure how common my experience is, but I do my thing in R and then just transfer the results over to Excel for others to use. I sympathize with the fact that R/SAS can be hard to learn. But there are many good resources that teach you how to do stuff, plus great recipe books like the O'Reilly series. $\endgroup$
    – rocinante
    Apr 2, 2014 at 23:54
  • $\begingroup$ I appreciate your virtuosity in all those complex programs such as R and SAS. But, is it Ok with you if I don't torture myself with those and use Eviews or STATA or even SPSS which seems a lot more user friendly than R and SAS. My objective is to become productive in a matter of days autonomously instead of years. I am dealing with time constraints. I have no argument regarding the commentary about Excel since I have been asked to upgrade as the mentioned methods in my question are not readily doable in Excel. $\endgroup$
    – Sympa
    Apr 3, 2014 at 0:43
  • 1
    $\begingroup$ I think learning new packages is part of the job with being a statistician. Every few years, for a variety of reasons I get to learn another. Each new thing I learn brings a few additional ways of thinking about a problem, or a different approach, or makes some kinds of analysis much easier than it was before. I don't think I've ever regretted the effort it takes. The more power to do good stuff you want, the more packages you need to be prepared to use. $\endgroup$
    – Glen_b
    Apr 3, 2014 at 2:42

2 Answers 2

Reset to default
1
$\begingroup$

Minitab is the easiest transition from Excel. It will do Generalized Least Squares but not ARCH/GARCH from the drop-down menu. If you poke around, you can see Youtube tutorials about how to do ARCH/GARCH in Excel.

This is what you're getting with Minitab in terms of the user interface. Whether this is worth the $1700 license is up to you.

enter image description here

enter image description here

enter image description here

Improve this answer
$\endgroup$
2
  • $\begingroup$ rocinante, thanks this is very helpful. The Minitab interface is the kind I love because I can actually do that. Gave you a +1 for that. $\endgroup$
    – Sympa
    Apr 3, 2014 at 17:34
  • $\begingroup$ @GaetanLion You are welcome. I would caution you against delaying learning programming because you're only making it worse for yourself, though. $\endgroup$
    – rocinante
    Apr 4, 2014 at 6:43
0
$\begingroup$

I have been using Gretl and IBM SPSS for analysis. Gretl is a great free software, but SPSS suits my needs much better. In my opinion, both are really user friendly. Since I am a Mac user, I am using a cloud version of the software.

Improve this answer
$\endgroup$
1
  • $\begingroup$ Thanks for your answer. Since I have asked this question a few years ago, I have acquired knowledge and proficiency in that stuff. And, I now use mainly EViews and R for econometric models and testing. I never heard of Gretl. Sounds interesting. $\endgroup$
    – Sympa
    Apr 21, 2017 at 0:28

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