free statistical software

[geneshackman]

Hi folks

I've been working on this article
http://en.citizendium.org/wiki/Free_statistical_software
still working....

Could I get feedback on how it's going so far?  Look useful? I have a reference in the article back to a website of mine, but I asked about this in the talk section, and it seemed to be okay.  Seem okay to you folks too?

thanks

Gene

[Hayford Peirce]

It looks pretty good to me, from the little I can follow it. I *did* do some minor editing to it yesterday, just syntax and stuff like that....

[Howard C. Berkowitz]

Hi folks

I've been working on this article
http://en.citizendium.org/wiki/Free_statistical_software
still working....

Could I get feedback on how it's going so far?  Look useful? I have a reference in the article back to a website of mine, but I asked about this in the talk section, and it seemed to be okay.  Seem okay to you folks too?

thanks

Gene

It's useful enough that I'm considering which packages to acquire (so many applications, so little time); of course, this makes me think about my assorted statistical needs. I need to think about this, but perhaps there might be some suggestions, or a table, which indexes packages by area of application.

The information about history, distribution, etc., is fine, but I suspect there would be utility to something that indicated which, for example, might be oriented to survey research vs. epidemiology (as two examples). If this were a table, it might also briefly mention the interface type: graphic/menu, freestanding language (I'm thinking SAS as an example), or something intended as a programming library.

[geneshackman]

Hayford wrote "It looks pretty good to me, from the little I can follow it. I *did* do some minor editing to it yesterday, just syntax and stuff like that.."

"from what little I can follow".  Does that mean the draft is not written to be easy to read?
Also, thanks for the editing!

Howard suggested a table of characteristics.  I'll try.  I think these are mostly general packages and can be used for any application. Most of the examples I found of published studies that used those packages were bio/chemical/public health, except easyreg which is economics.  I used them for social change study.

"graphic/menu, freestanding language "  I was thinking of adding something about that, I'll make sure I do.

Gene

[geneshackman]

Here is an idea I had.  I'll send an email to all the folks who created the packages and ask them a bunch of questions, saying it's for an article in CZ.  Then, I'll summarize their responses.  The questions would be along the lines that Howard was asking: are they aimed at any particular area of application, or are they more general, menu/command driven or both, range of procedures available, ability to import data, any other questions?

gene

[Howard C. Berkowitz]

Here is an idea I had.  I'll send an email to all the folks who created the packages and ask them a bunch of questions, saying it's for an article in CZ.  Then, I'll summarize their responses.  The questions would be along the lines that Howard was asking: are they aimed at any particular area of application, or are they more general, menu/command driven or both, range of procedures available, ability to import data, any other questions?

gene

Who knows? You might attract some here, or their students.

[Hayford Peirce]

No, no, it's easy enough to read, it's just that the whole concept is more or less beyond me, like most things since the invention of the quill pen....

[Howard C. Berkowitz]

No, no, it's easy enough to read, it's just that the whole concept is more or less beyond me, like most things since the invention of the quill pen....

If it helps, I have an abacus here as final backup.

[Joe Quick]

No, no, it's easy enough to read, it's just that the whole concept is more or less beyond me, like most things since the invention of the quill pen....

If it helps, I have an abacus here as final backup.

So if you use the beads on the abacus to physically create a curve with normal distribution, can you figure out the standard deviation by counting off beads?

[geneshackman]

Okay, I guess this article is pretty much close to done.  I'd like to ask if this could be a signed article?  Howard, would you be willing to nominate it?  Or is there a lot more to go on this article?  This might be one of a set of articles, one about commercial statistical software, one about statistical software in general, but this one on free packages, is the first.

thanks.  gene.

[Hayford Peirce]

I have a feeling that "signed article" really isn't the category here -- please read these guidelines carefully:

http://en.citizendium.org/wiki/CZ:Signed_articles

We've certainly discussed expanding the concept of signed articles (until the cows came home, actually, and then went to sleep), but nothing ever came of it.

I have a feeling that what you really want to do with this article is to start asking people (Editors) to look at it and to put it up for Approval, which is quite a different kettle of fish from signed articles....

But many other people here would know more about this than me....

[Howard C. Berkowitz]

I have a feeling that "signed article" really isn't the category here -- please read these guidelines carefully:

http://en.citizendium.org/wiki/CZ:Signed_articles

We've certainly discussed expanding the concept of signed articles (until the cows came home, actually, and then went to sleep), but nothing ever came of it.

I have a feeling that what you really want to do with this article is to start asking people (Editors) to look at it and to put it up for Approval, which is quite a different kettle of fish from signed articles....

But many other people here would know more about this than me....

I'm perfectly willing to nominate it, although as I've discussed with Gene, it may be worth having a couple of context-setting top level articles (e.g., statistical software). I'm comfortable with his expert observations on the free products, although if there is an issue of signing, the idea would be to separate the general aspects of free software (e.g., support or lack thereof) from his reviews.

[geneshackman]

Couple of questions:

1. If the article is not 'signed', can I still list it on my resume as an article that I wrote or for which I'm the primary author?  I'm also thinking of how to get others to contribute, and if they can put stuff on their resume, that certainly helps.

2. "separate the general aspects of free software (e.g., support or lack thereof) from his reviews."  Did you mean, from my reviews of individual packages?  Guess I'm not entirely clear on this.

3. "having a couple of context-setting top level articles"  Would I need to change the current article, or do you mean adding another article, more general, about statistical software?

4. "have a feeling that "signed article" really isn't the category here"   I'm not entirely clear about what doesn't fit the concept of signed article.  Does this article seem like it could be reference material?  I guess I'm not a world known expert on free statistical software, but then, I'm not too sure there is anyone else who is, except perhaps John Pezzullo, who has this page, like mine, http://statpages.org/javasta2.html    I haven't found any other reviews except those few listed in my review.  Or am I missing something about signed articles?

thanks

gene

[Hayford Peirce]

I really don't know much about signed articles, and I frankly find the subject very confusing.  However, I myself have a signed article.  It can be found at: http://en.citizendium.org/wiki/Matt_Helm/Signed_Articles

As you can see, by fiddling around, it is on a tab called Signed Articles belonging the article http://en.citizendium.org/wiki/Matt_Helm, so it is apparently a subpage of the Matt Helm article, NOT a stand-alone article by itself.

(I had nothing to do with its creation, except writing the article for an internet blog several years ago -- other people made a Signed Article out of it and put it where it now is.)

You will note that at the bottom of my signed article, there is a category link to other Signed Articles:

http://en.citizendium.org/wiki/Category:Signed_Articles

If you go there, you will find only about a dozen of them, including mine.  If you study those, perhaps that will give you a tip as to what you can do to make yours a signed article also....

Wish I could be more helpful!

[Howard C. Berkowitz]

Couple of questions:

1. Can you list articles on a resume?  This still hasn't been decided; signed or not.

2. By separating general aspect of free software, I'm not even sure this needs to be done. IF, however, there's something that clearly would be signed, it would be your reviews. Personally, I wouldn't object to having the general support issues in a signed article -- we just don't have a clear policy.

3. "Would I need to change the current article, or do you mean adding another article, more general, about statistical software?" Mostly adding; change the existing article only to link to it.

4. "have a feeling that "signed article" really isn't the category here"   Neither does anyone else! I do think the article is worth having, but it's a bit different than others, in a good way.