Opinion: Is Excel good for deep statistical analysis?

[otisvillain]

Hey all-

Do you think Excel is good for deep statistical analysis? Or would you use a specific statistical software instead?

 

[JeanKim]

Yes....
you can achieve everything Mini-Tab can do but you will need to build formulas or macros.
I have seen excel file that performs ANOVA and graphs and etc etc...

 

[Arachnyd]

I'm going to go with "NO"

I know that won't be popular on an excel forum, but you specifically said DEEP statistical analysis. Its certainly more than possible to use Excel, but it seems to make more sense to use something like "R". We've hit many-a-limitations in Excel which certainly can be worked around with complex formulas or macros (such as line limits for files and similar), but why not use something that can handle it with ease? It takes minutes to run something in "R" that may take hours in Excel, and with the price tag of "$0", there's not a high cost of entry.

Excel is not great for real complex data, and sometimes we use a combination of in-database processing via hadoop with R and some other analytical systems for the really "deep" stuff.

My data analysts often wonder why someone would use "Excel" when you can just use R for much of the same things... some of them rarely touch excel and do even basic table stuff in R.

 

[AhoyNC]

I would have to agree with Arachnyd. For "deep statistical analysis" a statistical software package like "R" would be the better choice. However "R" does have a steep learning curve.
"R" is also superior to Excel when it comes to graphing your statistical data.
SAS also has a free University edition (haven't tried so can't give an opinion on it), but here is a link.
Free Statistical Software, SAS University Edition | SAS

 

[63falcondude]

When I was in school for stats, we used minitab and then SPSS. Didn't get to R. It really depends on how deep is deep to you. I don't think Excel is the way to go here.