Help - Should I wait for STATA 15?

[Karthik Balachandran]

Dear all,
STATA noob here. Planning to buy STATA for academic use. Is it the right time to buy or should I wait for STATA 15? I hear its around the corner. Does anyone here know the approximate release date? If I buy now, will I get free upgrade? Sorry if this is not the right place to ask this.

[Richard Williams]

Stata 14 was released in April 2015. There is usually about a 2 year gap between releases. Personally I would buy 14 now.

[Justin Vung]

I can't confirm on personal subscriptions, but on my university computer (University of Alberta), I can type 'update all' and it will download a temporary copy of Stata 14.
(i.e. I can use Stata 14 commands, but this reverts to Stata 13 when I log out/shutdown the computer.)
My employer (Statistics Canada) runs Stata 14 on their computer, so I can't comment on their update policies.
I'm sure that a perpetual licence should include updates (I have free access now via VPN, so why buy?)

[Jorrit Gosens]

I can't confirm on personal subscriptions, but on my university computer (University of Alberta), I can type 'update all' and it will download a temporary copy of Stata 14.
(i.e. I can use Stata 14 commands, but this reverts to Stata 13 when I log out/shutdown the computer.)
My employer (Statistics Canada) runs Stata 14 on their computer, so I can't comment on their update policies.
I'm sure that a perpetual licence should include updates (I have free access now via VPN, so why buy?)

I can't replicate that on a personal license.

Licenses come with free updates, but not upgrades to newer versions. See here. Version 14 actually had a few nice features I would have liked to have as well, but it's not as if my current v13.1 is obsolete.

[Karthik Balachandran]

Thanks a lot for your guidance

[Richard Williams]

Every new version of Stata has some little enhancements I really really like. For example, Stata 14 improved both the margins command and the saveold command. And, of course, every new version has new statistical routines that I often don't use but which are no doubt somebody else's wildest dream come true.

Having said that, if I were condemned to only using Stata 12 or 13, I could probably survive ok.

The bigger question may be, does Stata 14 meet your current needs? If in doubt, you can request a 30 day evaluation copy. See

http://www.stata.com/customer-service/evaluate-stata/

[Phil Bromiley]

There are some procedures in Stata 14 that are much better than in 13. Look at the website for the big changes. From where I sit, the treatment model capability and the SEM capabilities both improved dramatically in 14.

[Sergiy Radyakin]

I hear its around the corner.

Depends on the source of the above information and the definition of "around the corner" in relation to your project.
If your academic year is September to May, then, perhaps, wait. If there is no seasonality and you plan to work intensively over the summer, then get your academic license now.

Best, Sergiy Radyakin

[Marcos Almeida]

Well, what about now, is it "around the corner", I mean, the launching of Stata 15?

[Richard Williams]

My records show that I have usually purchased new versions of Stata in June or July of odd-numbered years. I was therefore a bit surprised when Stata 14 came out earlier than that - I purchased it on April 22, 2015. So, based on past history, it might be unveiled tomorrow, or it might come out in the next 4 months. Only StataCorp knows for sure.

It is nice that, once purchased, you can just download it. It used to really annoy me when some places got access to physical disks a week or so before I did,

[Nick Cox]

See http://www.statalist.org/forums/foru...ed-ships-today for how the list learned about Stata 14.

An announcement from William Gould, President, is how people here learn about new releases. It's known that StataCorp works on an approximate 2 year cycle, but the exact date is a matter for StataCorp. All else is rumour and speculation. (I really don't know either.)

[Marcos Almeida]

Thank you Richard and Nick for the information. Since we are around the end of the 2 year cycle, the expectations rise!

[Nick Cox]

http://www.stata.com/support/faqs/re...tory-of-stata/ will allow confidence limits calculations too, and at the same time provide evidence against stationarity.